Can You Have a Family in Medicine?

Part of Before You Commit. The training window runs from about eighteen to somewhere in your thirties, and people build families at every point along it. This page lays out what the biology, the money, the schedule, and the Match actually ask of you, and it does not tell you what to do with any of it.

Why this page exists

A lot of people ask this question, and they tend to ask it quietly, as though wanting a family were a weakness in the plan. It is one of the more reasonable things to want to know before you spend your twenties on something.

So this is what I found when I went looking. Where the evidence is good the page says so, and where it is thin it says that instead of filling the gap.

The population-level numbers are solid: miscarriage by age, childcare prices, duty hours, the Match. Almost every figure specific to physicians comes from a voluntary survey, and people who have had a hard time answer those surveys more readily than people who have not. Those numbers show a shape without a reliable size.

The short version

Training runs from roughly eighteen to somewhere between thirty-one and thirty-eight, depending on the specialty. People have children before it, during it, and after it.

Three things press on that window and none of them press evenly. Fertility declines with age, and the decline is steeper for eggs than for sperm. Residency guarantees six weeks of paid parental leave once, and your co-residents cover it. In most of the country, center-based care for two young children costs more than a resident earns.

There is no verdict at the end of this page. Physicians raise four children and physicians spend five years in fertility treatment, and both are ordinary. What follows is the terrain, so that whatever you decide you decide knowing what it costs.

People do all of it

Research puts the share of trainees who become pregnant during residency or fellowship at 20% to 30%. Across specialties and survey methods the range runs from about 10% to 41%.1 Plenty of physicians have large families. Plenty have none, by choice, and are glad of it.

Not every family starts with a pregnancy. Adoption, fostering, surrogacy, and donor eggs, sperm or embryos are how a great many families begin, and for plenty of people they are a first choice rather than somewhere they arrived after a loss. This page does not cover them. It is built around one collision, the training clock against the fertility clock, and those routes have their own costs, waiting periods, home studies and approvals that the collision frame would handle badly. The omission is a gap in this site rather than a judgment about the routes.

Some try and cannot. That happens more often among physicians than in the general population, by an amount that is real and smaller than the internet will tell you.

Almost everything below is a constraint, and a page that only describes the hard end of a wide distribution is describing a tail.

Two clocks run at once

The first clock is the training itself, and it is arithmetic.
Median age starting medical school 23 (mean 24)2
95th / 99th percentile 28–30 / 33–362
Graduation about 27 to 283
Residency 3 years in family medicine, internal medicine and pediatrics, up to 7 in neurosurgery4
Fellowship 52.7% of subspecialty programs are one year; some run two or three5
First attending job 31 to 383

The second clock is fertility, and here the honest thing is to give you the society numbers and stop.

Chance of conceiving in a given cycle is roughly 20% at thirty and under 5% at forty, with overall fertility about half what it was by forty.6

Where those numbers come from matters, since this page is careful about that everywhere else. ACOG's 2025 statement on ovarian-factor fertility decline says that most of what is known about natural fertility rates comes from historical data, and the figure it prints to show the decline is built from populations like Hutterite marriages before 1921 and the Geneva bourgeoisie of the seventeenth century.6 The shape of the curve is not in question. The precision of any single point on it is softer than the way it usually gets quoted. Miscarriage rises alongside it: around 9% to 17% between twenty and thirty, about 20% at thirty-five, 40% at forty, 80% at forty-five.7 The largest population study of this, following 421,201 pregnancies, puts the low point at ages 25 to 29 (10%) and 53% at forty-five and over.8

Those two ladders disagree at the top end, and the disagreement is worth seeing. The bulletin's figures are clinical estimates carried through the literature; the registry study counted recorded pregnancies across a whole national population. Where they diverge, the registry is the better guide to how common something actually is.

One piece of proportion that usually gets left out of this argument. Everyone is delaying. The mean age at first birth in the United States moved from 26.6 to 27.5 between 2016 and 2023.9 The gap between physicians and everyone else is three or four years, not a decade.

What the numbers say, and what they have been made to say

Roughly a quarter of surveyed women physicians report an infertility diagnosis. The comparison group is where it gets slippery: depending on which federal figure you use, the general-population rate is somewhere between 8.5% and 19%.10

You will see this reported as "physicians are twice as likely." Sometimes it is. Sometimes it is 1.3 times as likely. The multiplier moves depending on which comparator the writer picked, and the ones that produce the biggest number tend to be the ones that get picked.

Every physician figure in this literature comes from a voluntary survey. There is no registry-based or claims-based study of physician fertility or obstetric outcomes. Not a weak one. None. That is a statement about the evidence rather than about you.

The specialty label predicts less than the hours behind it

The largest dataset on this, 2,519 physicians, found infertility in 26.8% of proceduralists against 26.5% of everyone else. The difference was not statistically significant.11 Surgery, by itself, did not show up.

What did show up, in a study of 692 female surgeons, was hours. Major pregnancy complications ran 48.3% against 27.2% in the comparison group, and the finding that transfers to a reader is this one: operating twelve or more hours a week in the third trimester carried an odds ratio of 1.57.12 The hours are measurable. The name of the field is a proxy for them.

Read the comparison in that study with its method in view. The women it compared surgeons against were the non-surgeon partners of surgeons, and their pregnancies were described by their surgeon husbands rather than by themselves. The hours finding sits inside the surgeon group and does not depend on that comparison, which is why it is the part quoted here.

This is also where "some specialties make it much harder" becomes something more useful than a rumor. A seven-year residency spends more of the window than a three-year one. A field whose hours run long spends more of your third trimester on your feet. A program where nobody has asked for leave recently is a different place to ask than one where three people did last year. Those are facts about schedules and about how people behave, and you can ask about all of them.

Freezing eggs: nobody publishes the price, and most people never use them

No professional society and no government body publishes a US price for elective egg freezing. Clinics publish prices, and this site does not take a price from the party selling the service. That leaves two anchors, and neither one is a quote.

The first is claims data. California's statutorily required review body priced mature oocyte cryopreservation at $10,078, a figure that covers retrieval, the drugs, the freezing itself and one year of storage.13 Treat it as an anchor: it is 2016 data, it reflects charges billed to insurers rather than a self-pay price, and it comes from medically indicated preservation rather than elective freezing.

The second is a survey of what clinics advertise. Across SART member clinics, IVF without medications averaged $9,279, and medications on their own averaged $4,398.13 Only 24.7% of clinics published any price at all. Most of the industry does not post one.

Three things hold regardless of which number you start from. Medications are usually quoted apart from the cycle, so an advertised price is rarely the whole price. Storage is an annual charge that runs for as long as the eggs are stored. And one retrieval is not always the whole course. Ask a clinic for the all-in figure in writing, including drugs, storage, and what a later thaw and transfer would cost, because that is the only way to get a number that applies to you.

Sperm banking is a different order of magnitude, and here the sources agree closely: roughly $350 to $470 to bank, and about $385 to $555 a year to store.14

In an analysis of the national registry, among women who froze eggs between 2014 and 2016, 5.7% came back to use them within five to seven years. Across everyone who froze, whether or not they returned, the live birth rate was 1.6%.15 The success rates in clinic marketing are calculated on the minority who come back.

The professional society is blunter about this than the industry is. ASRM's 2021 guideline states four separate times that there is insufficient evidence to counsel patients on the likelihood of a live birth after planned egg freezing, and warns about "a risk of false belief that future childbearing is 'guaranteed.'"16

For men the arithmetic is different. Banking sperm is cheap and requires no stimulation cycle. Advancing paternal age carries a real and modest effect, undetectable below forty, then a 1.23 relative risk at forty to forty-four and 1.43 at forty-five and over, adjusted for the mother's age.17 What is symmetric is the schedule. A co-parent in a surgical residency is absent whether or not they gave birth.

Six weeks, and what the six weeks costs the people around you

For comparison, since most people entering medicine have never seen a corporate leave policy: Microsoft gives twenty weeks to a birthing parent and twelve to any new parent. Salesforce gives twenty-six weeks to a primary caregiver and twelve to a secondary one, at full base salary.18

In residency the guaranteed number is six weeks at 100% of salary, once, at any point in training.19 Taking it has two consequences a corporate leave does not.

Your colleagues cover it. In a survey of neurology program directors, answered by a third of them, 100% said other residents covered parental leave. Attendings covered at 14% of programs. The most-cited barrier to offering more than six weeks paid was "equity concerns and impact on other trainees," at 82%, ahead of staffing at 80%.20

Past a point, it delays you. Six weeks will not push back your board eligibility. Beyond what your own board allows, the time gets made up at the end. The boards vary more than a single number suggests: emergency medicine and surgery allow eight weeks, anesthesiology and obstetrics and several others twelve, family medicine up to twenty across all of training, internal medicine ten to twenty depending on program length.21

Pediatrics is the one to understand properly. Its additional eight weeks is a waiver the board may grant, if the program director and the Clinical Competency Committee agree, considered in the final three months of training.22 That is not something you can plan around in advance.

Both of those facts are true at once. The redistribution onto your co-residents is why the guarantee stops where it does, and you will spend years on the covering side of it yourself.

If you are weighing this against a partner's job in another field, the difference in leave above is the smallest part of it. The schedule is the rest.

What is guaranteed, what is only disclosed, and what is neither
Guaranteed Six weeks of approved medical, parental and caregiver leave at 100% of salary, once, at any point, starting the day you are required to report, plus a further week of paid time off and continuation of health and disability insurance.19 Six weeks will not delay board eligibility.21
Only disclosed Programs must tell applicants they interview what their leave policy is. They are not required to publish it. Two years after the requirement took effect, 37.5% of emergency medicine residency programs had nothing about parental leave or pregnancy accommodation on either their program or their GME website.23
Neither Federal job protection during intern year.24 Fertility coverage, which no ACGME, ABMS or AAMC standard requires. On-site or backup childcare. Who covers your call.

Two points of precision, because most coverage of this gets both wrong. The leave requirement sits in the ACGME's Institutional Requirements rather than the Common Program Requirements, which means it binds the sponsoring institution rather than your program.19 And the board policy is a guarantee about board eligibility, not about employment. It says nothing about your pay, or about whether your program grants the leave in the first place.21

Then the finding that outlasts any policy change. In a study of family medicine programs before the mandate, residents and faculty both took significantly less leave than they were offered, and how much leave people took correlated within programs.25 The binding constraint was culture. A written policy and a lived norm are different objects, and finding out about the second is most of what a program visit is for.

Daycare costs more than your rent, and you are on a resident's salary

A first-year resident averages $68,166, reaching about $73,000 by the third year.26 Center-based care for one infant ran between $7,700 and $26,300 a year depending on the state in 2024, and above $31,000 in the most expensive counties. Two children in a center, one infant and one four-year-old, ran from $14,560 in Mississippi to $47,174 in Washington, DC.27

Against a first-year stipend, that is 21% to 69% of pre-tax income.28 Before tax. Before rent. Before loans sitting in forbearance and growing.

The federal standard for affordable childcare is 7% of household income. One study measured every sponsoring institution in the country against it. Childcare failed at 91.6% of them.29 Its median came in at 10.3%, well under the arithmetic above, because it priced one infant instead of two children and counted a partner's income alongside the resident's. Only 11.9% of institutions offered a subsidy. Only 25.5% offered care on site.

Between 2000 and 2023, childcare costs rose 26.4% after inflation. Resident salaries rose 1.2%.

Your own number sits between those two pictures, and which one it sits nearer comes down to whether there is a second income and how many children are under five at once.

This is also where the advice to space your children stops being a preference and becomes arithmetic. The expensive infant band ends before the second birthday, but the price steps down only 10% to 20%. The real drop is kindergarten at five, and only 37% of four-year-olds are in state-funded pre-K.30 A resident with two children under five pays close to two full prices for years.

Spacing them so the first is in school before the second arrives is a rational response to that. It also runs straight into the fertility timeline three sections up. Both things are true and this page does not resolve them.

Then the parts that are not about money.

The hours do not line up. Duty hours are eighty a week averaged over four weeks, with shifts up to twenty-four hours plus four, and in-house call as often as every third night.31 Centers are not open for that. Care during overnights and weekends exists mostly through home-based providers rather than centers, and the federal agencies that track this describe the supply data as an acknowledged gap.32

A sick child has nowhere to go. Centers exclude children for fever, and the familiar rule, fever-free for twenty-four hours without medication, is a state or center convention rather than a national standard. The national guidance is narrower: it calls for a child to be sent home at 100.4°F or above when there is a behavior change or another symptom. Under two months of age, a fever at that threshold means exclusion and immediate medical attention whether or not anything else is wrong.33 In practice your center's rule is the one that governs. A resident who is post-call with an excluded child has no slack anywhere in the system, and parents dose a fever down and send the child anyway, because the alternative is a shift that nobody is free to cover.

Lactation. In a national survey of resident physicians, answered by 15.6% of those invited, 91% had a lactation room and 21% called it usable.34 Sixty percent lacked adequate milk storage. Thirty-seven percent stopped lactating before their own goal. Forty percent said faculty, and forty percent said co-residents, had made them feel guilty about pumping.

And one thing that worked. A randomized trial across seven institutions gave 143 pregnant and postpartum trainees a support package: a wearable pump, a smart bassinet, virtual perinatal support, and formal faculty mentorship. Burnout held flat in the group that got it and rose in the group that did not. The package cost about $2,300 a person, against an estimated $7,600 per-physician cost of burnout.35 It is the only interventional evidence on this page.

The Match decides where you live, and it does it three times

Medical school admissions, the Match, and then fellowship. Three placements across about a decade, and none of them are yours to choose outright. This is where a partner's career, aging parents, and whatever family support you have all collide, and it is the constraint least amenable to planning.

In the 2026 Match there were 1,258 couples. Both partners matched in 1,113 of them, 88.5%. In 114 couples only one matched, and in 31 neither did.36

You will see 93.0% quoted for couples. That is an individual-level rate and it sits essentially on top of the overall rate for everyone. The number that describes your risk of being separated is 11.5% of couples.

One piece of mechanics, before you rank. The two lists have to be the same length, the algorithm matches pairs of programs rather than individuals, and if a couple does not match as a couple the algorithm does not fall back to processing the two lists separately.36

The FLI version: no cushion under any of it

Everything above is harder without money behind you, and the ways are specific.

Fertility preservation is a cash expense. There is no loan product for it, and it arrives exactly when your net worth is at its most negative, and often when other people are relying on your future income.

Childcare is the same problem arriving monthly instead of once. The resident who can ask a parent to take a Tuesday afternoon pays nothing for it. The resident who cannot pays for every hour.

Insurance mandates help less than the map suggests. Sixty-one percent of covered workers are in self-funded plans, which are federally regulated and exempt from state insurance mandates, and large academic medical centers commonly self-fund. So training in a state with a mandate guarantees you nothing. You will also see two different counts of how many states have one: fifteen require coverage and eleven include IVF, while another count reaches twenty-five plus DC by also counting states that only require insurers to offer a plan. For someone asking what they will actually receive, the smaller number is the relevant one.37 No state Medicaid program covers IUI, IVF, or cryopreservation.38

What money buys here, named plainly: taking the six weeks without an unpaid tail on the end, backup care on the day a center sends your child home, a flight home when something happens, a research year you can absorb. Someone without it makes the same decisions with fewer of those options on the table.

What this page cannot tell you

None of this is medical advice, and it is not a substitute for a conversation with a reproductive endocrinologist. The one genuinely actionable line is the threshold for having that conversation: twelve months of trying if you are under thirty-five, six months at thirty-five and over.39

Parenting during residency, as opposed to deciding about it, belongs to Medical School. The loan mechanics belong to The Money. And whether you want any of this is the input that every number on this page depends on, which is the one thing a page cannot supply.

Where this fits

Is Medicine the Right Path? is the broader version of this question, and The Friend Who Went Into Tech runs the same decade as a ledger, which is the half this page leaves out. If a leave or a pause is already on your mind, Taking a Leave covers the mechanics inside medical school and Nontraditional Paths covers arriving later than the median. Accommodations Before Day One is where disclosure and accommodation get handled.

This page is educational and reflects sources current as of its review date; it isn't medical, financial, or legal advice. Last reviewed: 2026-08-12.

References

ACOG Committee Statement No. 22, Anticipatory Counseling Regarding Ovarian-Factor Fertility Decline, Obstetrics & Gynecology, November 2025, 146(5):e98–e104 replaces Committee Opinion No. 589 and is the current statement on this subject. It was read in full for this review and it publishes no per-cycle probability ladder by age, so it neither confirms nor supersedes the per-cycle figures above. What it does say is the sentence quoted in the body: "The majority of understanding of natural fertility rates comes from historical data." Its Figure 1 plots marital fertility rates for ten historical populations, among them Hutterite marriages before 1921, the Geneva bourgeoisie with husbands born 1600–1649, Canadian marriages of 1700–1730, and Normandy marriages of 1760–90. Its own quantitative anchor is a 1982 donor-insemination study in which pregnancy over twelve months of cycles fell from 74% under age 31, to 62% at 31–35, to 54% over 35.

One finding from it worth carrying into any conversation you have about this: the statement reports that obstetrics and gynecology residents were no more likely than residents in other specialties to answer fertility and oocyte-preservation knowledge questions correctly.

That 290 comes from the paper's correction notice, which raised the count of surgeons who experienced a pregnancy loss from 90 to 290 and states that "all other information in the article was correct and is unchanged." No figure used on this page is affected, and the open mirror carries the corrected number.

Two things this paper is not a source for. It reports assisted reproductive technology use, 172 of 692 surgeons, rather than a diagnosed infertility rate. And the "32% of female surgeons experience infertility" figure widely attributed to it appears nowhere in it.

The first is the California Health Benefits Review Program, the University of California body statutorily required to analyze insurance mandate bills, which priced mature oocyte cryopreservation at $10,078, stating that the unit cost covers "medical, surgical, and drug costs related to retrieving and preserving sperm, oocytes, and embryos, culturing of oocytes and embryos, and storage for 1 year." Three limits are stated in the report and repeated in the body: the estimate uses 2016 CHSD and MarketScan claims with Milliman's 2016 cost data; it reflects charges to carriers rather than self-pay prices; and it sits in an analysis of fertility preservation for enrollees with cancer, so it is medically indicated preservation rather than elective freezing. The same table gives embryo cryopreservation at $11,254 and sperm at $468.

The second is Larsen et al., Translational Andrology and Urology 2020, which surveyed SART clinic websites and found IVF without medications at a mean of $9,279 and medications alone at $4,398. Note that this is IVF rather than oocyte cryopreservation, which is why it is presented as a second anchor rather than as the price of egg freezing. That survey also found that only 24.7% of clinics published any price at all, and it reports no geographic variation, finding no significant difference even between states with and without an insurance mandate.

A note for anyone re-checking this page. Until 2026-08-15 the body carried retrieval at $8,500–$11,000, medications at $3,000–$6,000 on top, a $12,000–$16,000 realistic total rising to $18,000–$20,000 in New York or the Bay Area, storage at $500–$1,200 a year, a $30,000–$40,000 planning figure, and $4,000–$12,000 to thaw and transfer. None of those traced to either source cited here, the geographic and thaw figures appear in neither, and stacking a medication cost onto the CHBRP figure double-counted drugs it already included. They were removed rather than re-sourced, because the only place those numbers exist is clinic marketing.

Footnotes

  1. AAMC, "When pregnancy and residency coincide", 17 June 2026, for the 20–30% figure. The AAMC does not name the underlying study, so the attribution here is to the AAMC. The wider 10%–41% range reflects variation across specialty-specific surveys and survey methods. A commonly quoted "40% of residents are parents" traces to a single institution-system survey of about 650 residents in 2016 and is not used here. ↩

  2. AAMC FACTS Table A-6, Age of Applicants to U.S. Medical Schools at Anticipated Matriculation by Sex and Race/Ethnicity, 2014-2015 through 2017-2018, read from the Wayback snapshot of 12 July 2025. Figures are the matriculated column: median 23 for both women and men across all four years; 95th percentile 28 for women and 29–30 for men; 99th percentile 33–34 for women and 35–36 for men. ⟳ Verify: this table has been discontinued and the 2025 FACTS release contains no age table, so the figure is dated to the 2017-18 entering class rather than presented as current. Corrected 2026-08-15: the live AAMC URL now 404s and is replaced with the archived snapshot, matching how this site handled AAMC's retired active-physicians table. The percentile ranges previously read 28–29 and 33–35, both narrower at the top than the table's matriculated columns. ↩ ↩2

  3. Arithmetic rather than a citation. Graduation age is the median matriculation age plus four years. The first attending job is that sum plus residency length from requirement 4.1 and, where applicable, a fellowship. Both are shown as ranges because the inputs are ranges. ↩ ↩2

  4. ACGME Program Requirements, requirement 4.1, 2025 reformatted, read per specialty. Family medicine: "The educational program in family medicine must be 36 months in length." Neurological surgery: "must be 84 months in length." Internal medicine and pediatrics are likewise 36 months. Corrected 2026-08-15: this footnote previously linked the requirements directory, which is not browsable and returns 404; it now links the two specialty documents the claim rests on, both quoted. ↩

  5. ACGME Data Resource Book 2024-25: 52.7% of subspecialty programs are one year in length. That share is the figure the Data Resource Book reports; the two- and three-year lengths are ordinary field norms rather than a second number taken from it. ↩

  6. ASRM, Age and Fertility patient booklet. The booklet was created in 2012, so the "about half by forty" figure is carried here on the strength of ASRM's Optimizing Natural Fertility guideline instead, which states that fertility "is decreased by about half at age 40 compared with women in their late 20s and early 30s." ↩ ↩2

  7. ACOG Practice Bulletin 200, Early Pregnancy Loss. ↩

  8. Magnus et al., BMJ 2019, a population-based study of 421,201 pregnancies. ↩

  9. CDC/NCHS, National Vital Statistics Reports 74(9), mean age of mothers, 2016–2023. A figure of 23.6 years for the general population circulates in this discussion and is contradicted by this source; using it roughly doubles the apparent gap. ↩

  10. Stentz et al., Journal of Women's Health 2016 is the origin of the roughly one-in-four figure: n=327, 54.5% response rate, with a denominator of women who had attempted conception. The general-population comparators are CDC/NCHS National Health Statistics Report 202 (8.5% and 13.4% on different definitions) and the CDC infertility FAQ (19%). The choice of comparator is what moves the multiplier from about 1.3 to about 2.8, which is why no multiplier appears in the body. A 2023 Annals of Surgery study of 4,533 respondents recruited entirely through private female-physician social media groups, with no computable response rate, is the most likely source of upward bias in this literature and is not used here. ↩

  11. Kelly et al., Journal of Assisted Reproduction and Genetics 2025, n=2,519: infertility 26.8% in proceduralists against 26.5% in non-proceduralists, p=0.66. ↩

  12. Rangel et al., JAMA Surgery 2021 (open mirror), n=692 female surgeons: major pregnancy complications 48.3% against 27.2%, adjusted odds ratio 1.72 (1.11–2.66); operating twelve or more hours weekly in the third trimester, odds ratio 1.57 (1.08–2.26). Two cautions. The comparison group was the female non-surgeon partners of surgeons, described by their surgeon husbands. And the widely quoted 42% figure is a per-person lifetime rate rather than a per-pregnancy one, which is why it does not appear in the body: 290 of the 692 surgeons (42.0%) had ever had a pregnancy loss, which is not the same kind of number as the per-pregnancy rates in the miscarriage ladder earlier on this page, and the two should never be set against each other. ↩

  13. No professional society or government body publishes a US price for egg freezing, and ASRM's ethics opinion says only that it is expensive and usually self-pay. Clinic and fertility-finance websites are the only readily available source of dollar figures, and this site does not treat a seller as a price source. Only the two figures named in the body are used, and no total is derived from them, because the arithmetic that would produce one — cycles per course, eggs per retrieval, a thaw-and-transfer price — has no non-seller source. ↩ ↩2

  14. Three sources cluster tightly here, which is why this one gets a range and egg freezing does not. Su et al., Andrologia 2021: median sperm bank fee $350 and median annual storage $385, with under a fifth of registered facilities publishing prices. Larsen et al. 2020 found sperm cryopreservation at $388 and annual storage at $555, though off only four reporting clinics. The CHBRP claims analysis puts sperm cryopreservation at $468 inclusive of one year of storage. The body's $350–$470 and $385–$555 are the spans of those published figures rather than a derived estimate. ↩

  15. UCLA analysis of the national SART registry, American Journal of Obstetrics and Gynecology, August 2025, reported here: among women who froze eggs in 2014–2016, 5.7% returned to use them within five to seven years, and the live birth rate across everyone who froze was 1.6%. ↩

  16. ASRM, evidence-based outcomes after oocyte cryopreservation, 2021. The guideline states in four separate places that there is insufficient evidence to counsel patients on live birth likelihood after planned cryopreservation, and warns of "a risk of false belief that future childbearing is 'guaranteed.'" ↩

  17. du Fossé et al., Human Reproduction Update 2020. No detectable effect below forty; 1.23 (1.06–1.43) at forty to forty-four; 1.43 (1.13–1.81) at forty-five and over, adjusted for maternal age. ↩

  18. Microsoft leave of absence benefits and Salesforce parental benefits, April 2026. These two are used because both publish current policies directly; figures circulating for other large technology employers are older reporting or disputed in practice. ↩

  19. ACGME Institutional Requirements, effective 1 July 2022 and enforced by the Institutional Review Committee from 1 July 2023, at IV.H.1 — renumbered to 4.8 on 3 September 2025. Both numbers are given because this requirement has already been renumbered once. Note that this sits in the Institutional Requirements rather than the Common Program Requirements, which contain no numbered requirement mentioning leave, parental leave, or caregiver leave. ↩ ↩2 ↩3

  20. Conway et al., BMC Medical Education 2024, a survey of neurology program directors with a 33% response rate: 100% reported co-resident coverage of parental leave, 14% attending coverage, and 82% cited equity concerns and impact on other trainees as a barrier to offering more paid leave. ↩

  21. ABMS Policy 3.02.12, effective 1 July 2021, and the member board comparison. The policy guarantees that six weeks will not delay board eligibility. It is a board-eligibility guarantee and not an employment one: it says nothing about pay or about whether a program grants the leave. ↩ ↩2 ↩3

  22. American Board of Pediatrics, absences from training policy. The additional eight weeks is a waiver the board may grant, requiring agreement from the program director and the Clinical Competency Committee, and considered in the final three months of training. ↩

  23. Hoag et al., Cureus 2026, website data collected July 2024: 37.5% of emergency medicine residency programs published no parental-leave or pregnancy-accommodation information on either their program or their GME website. ↩

  24. US Department of Labor Fact Sheet #28, revised March 2025. FMLA eligibility requires twelve months of tenure with the employer and 1,250 hours worked. At residency hours the hours test is cleared in roughly four months, so tenure is the binding constraint, and an intern is generally not eligible until about July of their second year. Prior employers do not count toward tenure, and transferring institutions restarts the clock. This is arithmetic from the rule rather than a finding in a study. ↩

  25. Wendling et al., Family Medicine 2019, a pre-mandate baseline in family medicine: residents and faculty both took significantly less leave than their programs offered, and leave-taking correlated within programs. ↩

  26. AAMC Survey of Resident/Fellow Stipends and Benefits, 2025, as reported by the AMA. These are unweighted averages rather than medians. ↩

  27. State-level and county-level prices from the US Department of Labor National Database of Childcare Prices. The two-child figures ($14,560 in Mississippi to $47,174 in DC) come from Child Care Aware of America's 2024 affordability analysis; Child Care Aware of America is an advocacy organization and is identified as one wherever its figures are used. ↩

  28. Arithmetic rather than a citation: published two-child center costs against the published first-year stipend, both pre-tax. ↩

  29. Brewster et al., "Childcare Affordability and Benefits Among Resident Physicians," JAMA Network Open, 16 May 2025, doi:10.1001/jamanetworkopen.2025.11089, open access. Two things about the method matter for reading the numbers. The unit is the sponsoring institution (n=936), not the residency program. And the affordability index assumes a two-partner household with a single infant in center-based non-residential care, priced from 2023 county-level Department of Labor data against county median income from the Census. It is therefore the optimistic case; a single-parent resident, or one with two children, is worse off than the index shows, and the authors name the inability to adjust for family arrangement as a limitation. The 7% threshold is the Department of Health and Human Services definition the study applied; that benchmark was a voluntary standard derived from 2011 Census data, made binding by the 2024 Child Care and Development Fund rule and repealed effective 2026, so it is described here as the standard that study used. A note for anyone re-checking this page: the widely circulated version of this study, including its institutional release, reports 935 programs, a 12.7% median, 98.2% unaffordable, 25.6% on-site, and childcare costs rising 44.7% against resident salaries falling 0.23%. Five of those six do not match the published paper, and the salary figure is reversed in direction. The figures above are read from the article itself. ↩

  30. NIEER, State of Preschool Yearbook 2023-24: 37% of four-year-olds are enrolled in state-funded pre-K. ↩

  31. ACGME Common Program Requirements (Residency), effective 1 July 2026, section VI.F: eighty hours per week averaged over four weeks, clinical work periods not exceeding twenty-four hours with up to four additional hours for transitions of care, and in-house call no more frequent than every third night averaged over four weeks. ↩

  32. HHS Administration for Children and Families on nontraditional-hour care, and Urban Institute analysis of the 2019 National Survey of Early Care and Education for demand. No national supply-side count of centers offering overnight or weekend care exists, which is why none is given here. ↩

  33. Caring for Our Children Basics, Standard 3.6.1.1, Inclusion, Exclusion, and Dismissal of Children, hosted by the HHS Administration for Children and Families, last reviewed 6 May 2026. The standard directs that a child be sent home at 100.4°F or above with behavior change, and lists "causes a fever with behavior change or other symptoms" among its exclusion criteria. For infants younger than two months a temperature at that threshold requires exclusion and immediate medical attention with or without any other symptom. It contains no fever-free-for-24-hours rule, and it states that most conditions requiring exclusion do not require a doctor's visit before the child returns. That 24-hour rule is a state and center convention; South Carolina's exclusion list is one example of it. Corrected 2026-08-15: this footnote previously cited two nrckids.org URLs, both now dead, and stated the age threshold as four months. The current standard says two. ↩

  34. Peters et al., BMC Pregnancy and Childbirth 2020, a national survey of resident physicians with a 15.6% response rate. ↩

  35. Rubio-Chavez, Rangel et al., JAMA 2026, a randomized trial across seven institutions in 143 pregnant and postpartum trainees. Corrected 2026-08-21: an institutional summary was cited alongside the trial and is removed — Mass General Brigham moved its newsroom without preserving the URL, and it now redirects to the site root with a browser user agent as well as an automated one. The JAMA article it summarized is the source and is unaffected. ↩

  36. NRMP, Results and Data: 2026 Main Residency Match: 1,258 couples, 1,113 with both partners matched (88.5%), 114 with one matched, 31 with neither. The 93.0% figure commonly quoted is an individual-level rate and sits essentially on top of the 93.3% overall PGY-1 rate. Mechanics from NRMP, Couples in the Match. ↩ ↩2

  37. KFF counts fifteen states requiring insurers to cover infertility treatment, eleven of which include IVF; RESOLVE reaches twenty-five states plus DC by also counting states that require insurers to offer a plan. The counts differ because they are counting different obligations. Self-funded employer plans are regulated federally under ERISA and are exempt from state insurance mandates, and 61% of covered workers are in one. ↩

  38. No state Medicaid program covers intrauterine insemination, IVF, or cryopreservation. ↩

  39. ASRM, Optimizing Natural Fertility, 2022: evaluation is recommended after twelve months of attempting conception under age thirty-five, and earlier evaluation "is warranted after 6 months without conception for women aged 35 years and older due to the age-related decline in fertility." The two societies word the boundary differently at exactly thirty-five, which matters only if that is your age: ASRM puts thirty-five in the six-month group, while ACOG Committee Statement No. 22 summarizes the same criteria as one year for "35 years or younger" and six months for "36–40 years." The threshold in the body follows ASRM's own wording. At thirty-five, ask rather than work it out from a page. ↩