Alcohol Before IVF: How Long Should You Stop, and Does He Have To?

Evidence-based · Written by Dr. Leila Fazlicic, D.Ac, L.Ac · All key claims cited to peer-reviewed research

Behind that sit three separate questions almost every article merges. Does alcohol reduce your chance of conceiving? Weak and inconsistent — the largest cohort found nothing below 14 drinks a week. Does it reduce IVF success? Two good cohorts flatly disagree, and the best meta-analysis returned a maternal effect that was not statistically significant. Is it safe once you might be pregnant? No, and that is not disputed. For the male partner, the signal is real but concentrated in heavy and daily drinking, not in a glass of wine on a Friday.

Working out what this means for the two of you specifically is the hard part to do alone.

Book a free 15-minute call · See what working together looks like

Before anything else: if you had wine at a wedding four months before your last retrieval and you have been turning that over ever since, put it down. Nothing in the literature supports the idea that a social drink decided your cycle. If nobody told you this before, that is not your failure — the dose evidence rarely makes it onto a clinic handout.

How long before IVF should you stop drinking?

Nobody knows, and it is worth saying that plainly, because the confident numbers you will find online are not derived from studies that tested them.

Here is what actually anchors the common advice. The largest IVF cohort measured alcohol intake at the start of the cycle and nowhere else (Rossi et al., 2011). An earlier, smaller study of 221 couples found associations across several different windows — fewer eggs retrieved for intake in the year before the attempt, higher odds of not conceiving for the month before, and higher miscarriage risk for the week before (Klonoff-Cohen et al., 2003). Its confidence intervals are very wide and its headline figures should not be quoted as reliable. And on the male side, a stable-isotope study in 11 men established that a sperm ejaculated today began developing a mean of 64 days ago, with a range of 42 to 76 (Misell et al., 2006).

That last number is the only hard timeline in the whole conversation. It does not say alcohol harms sperm. It says that whatever he changes, it takes about two to two and a half months to show up in a sample. Which is why "about three months before the cycle for him" is a reasonable construction — and why anything shorter cannot be expected to register. The same arithmetic underlies every other male-side input, including the ones with better evidence behind them.

Count back 3.5 months from your next cycle. That is when the window opens, and that is all the biology asks. Not more.

Does alcohol reduce IVF success rates?

This is where the field is genuinely split, and you should see both sides.

The study that started it

A prospective cohort of 2,545 couples across 4,729 IVF cycles in Boston found that women drinking four or more drinks per week had a live birth odds ratio of 0.84, and couples where both partners drank four or more had an odds ratio of 0.79. Failed fertilization was more common in both partners' higher-intake groups (Rossi et al., 2011).

Two honest caveats. Alcohol was assessed only at the start of the cycle, so nothing is known about intake during stimulation or after transfer. And the paper's beverage-specific findings — white wine implicated in women, beer in men, red wine in sperm concentration — form a pattern with no plausible mechanism and are best read as the product of testing many comparisons at once.

The study that complicates it

A Danish cohort of 1,708 women followed through 1,511 IUI, 2,870 IVF/ICSI and 1,355 frozen transfer cycles, with complete national registry follow-up — meaning nobody was lost to follow-up — found nothing. Live birth after IVF or ICSI relative to abstainers: 1–2 drinks a week, relative risk 1.00. Three to seven, 0.95. More than seven, 0.89, with a confidence interval from 0.53 to 1.51. Binge episodes in the month before baseline were unrelated to live birth (Lyngsø et al., 2019).

Their conclusion, in their own words, is that "it is not necessary to abstain from alcohol when striving for a successful outcome following fertility treatment."

What the pooled evidence says

A 2022 dose-response meta-analysis covering 26,922 women and their partners undergoing IVF or ICSI found a maternal alcohol association with pregnancy of odds ratio 0.83 — with a confidence interval of 0.69 to 1.01, which crosses the null. The paternal association with live birth was odds ratio 0.88 (0.79–0.99), just clearing significance. Modelled dose-response put the reduction at roughly 7% for women and 9% for men at 84 grams of ethanol per week, about six US standard drinks (Rao et al., 2022).

So the honest summary is this: the female IVF signal is weak, inconsistent, and not statistically robust when pooled. The male signal is modest but survives. Neither justifies the certainty with which this advice is usually delivered.

Does he need to stop drinking too?

Yes — and the case is narrower than you have probably been told. It is about heavy and daily drinking, not about any drinking.

The best-designed synthesis of semen quality pooled 15 studies and 16,395 men. It found a detrimental association with semen volume of 0.25 mL and with normal morphology of 1.87 percentage points. Crucially, the difference was more marked comparing occasional versus daily drinkers than never versus occasional — which the authors read as moderate consumption not adversely affecting semen parameters (Ricci et al., 2017). Their own recommendation stops at "avoid heavy alcohol drinking."

A 40-study meta-analysis of 23,258 men found the same shape more explicitly: no change in any semen index in the moderate group (under 7 units a week), and harm to semen indices and sex hormones in the heavy group (over 7 units a week). Total testosterone was substantially lower in drinkers; semen volume was reduced; and notably, there was no effect on sperm DNA fragmentation across the pooled studies (Nguyen-Thanh et al., 2023).

The clearest dose data comes from 1,221 young Danish men examined in a general population setting rather than a fertility clinic. Sperm concentration, total count and normal morphology all fell with rising habitual intake, detectable from about 5 units a week and most pronounced above 25. Men with a typical intake above 40 units a week — roughly 34 US standard drinks — had a 33% lower sperm concentration than men drinking 1 to 5 units (Jensen et al., 2014). That 33% figure circulates widely without its context. The context is the whole finding.

The finding worth knowing if he flushes

Roughly 40 to 50% of people of East Asian descent carry the ALDH2*2 variant — the one responsible for facial flushing — which impairs clearance of acetaldehyde, alcohol's first metabolite. In 112 East Asian men, carriers of the variant who drank had markedly lower total motility than carriers who did not drink — median 20% versus 43% — and lower motility than drinkers without the variant, 20% versus 41% (Greenberg et al., 2022).

This is a small, single-site, cross-sectional study, so hold it loosely. But it is the cleanest human evidence that acetaldehyde rather than ethanol itself is the proximate concern — and it identifies a group for whom the general reassurance about moderate drinking may simply not apply. If he flushes when he drinks, that is worth mentioning to his physician, and worth factoring into the rest of his 74-day plan.

What about trying to conceive naturally?

The two largest cohorts point in reassuring and slightly awkward directions.

Among 6,120 Danish women trying to conceive without treatment, adjusted fecundability ratios were 0.97 at 1–3 servings a week, 1.01 at 4–7, 1.01 at 8–13, and 0.82 at 14 or more — every estimate crossing the null. The authors concluded that consumption below 14 servings a week "seemed to have no discernible effect on fertility" (Mikkelsen et al., 2016).

For men, a pooled analysis of 2,679 couples across Danish and North American preconception cohorts found no association at any intake level — 1–5 servings a week, fecundability ratio 1.02; 6–13, 1.10; 14 or more, 0.98. The two cohorts disagreed with each other in the middle categories, which the authors flag directly (Høyer et al., 2020).

One study does support a timing message. Using daily diaries rather than recalled weekly averages, a cohort of 413 women found reduced fecundability for moderate and heavy drinking in the luteal phase, and a 41% reduction per additional binge day in the ovulatory sub-phase (Anwar et al., 2021). It is one modest cohort with data collected in the early 1990s, and two other well-powered studies found no independent binge effect. Take it as a reason to skip the heavy night around ovulation, not as a rule.

A drink is not a drink: the unit problem

Half the confusion in this topic is arithmetic. The studies do not agree on what one drink is, and neither do the guidelines.

What one "drink" means in each source, and the threshold each one uses

Source	One "drink" or "unit"	Threshold used	In US standard drinks (14 g)
Rossi 2011 (US IVF cohort)	~12–14 g ethanol	4 drinks/week ≈ 50 g	≈ 3.5/week
Rao 2022 (meta-analysis)	grams	84 g/week	6/week
Jensen 2014 (Danish men)	Danish unit = 12 g	Signal from 5 units; pronounced above 25; 40 units = 480 g	≈ 4/week; ≈ 21/week; ≈ 34/week
ASRM 2022	10 g ethanol	Concern begins above 2 drinks/day = 20 g/day	≈ 10/week
NICE (UK)	UK unit = 8 g	Women 1–2 units once or twice weekly; men 14 units/week spread out	≈ 0.5–1/week; ≈ 8/week
CDC (US)	US standard drink = 14 g	Binge = 4 (women) / 5 (men) per occasion	4 / 5 per occasion

A "drink" in the Rossi study is about 1.75 UK units. Its four-drinks-a-week threshold is roughly seven UK units — half of what NICE tells British men is unlikely to affect semen quality. Any advice quoted without its unit system is not advice; it is noise.

What the guidelines actually say — and why clinics say something stricter

This surprises most couples. The professional bodies are considerably more permissive than the instruction sheet from the clinic.

The American Society for Reproductive Medicine's 2022 committee opinion states that higher levels of alcohol consumption by women — more than two drinks per day — "probably are best avoided when attempting pregnancy, but there is limited evidence to indicate that more moderate alcohol consumption adversely affects fertility." On men, it states that "a dose–response pattern has not been established, and there is a lack of evidence for any effect of moderate alcohol consumption on male fertility" (ASRM Practice Committee, 2022).

The UK's NICE guideline NG257 tells men that drinking within 14 units a week, spread across several days, "is unlikely to affect their semen quality," and advises women trying to conceive to keep to no more than 1 or 2 units once or twice a week while avoiding intoxication.

Neither body grounds an abstinence recommendation in conception odds. NICE does tell clinicians to advise women that, according to the UK Chief Medical Officer's guidance, the safest approach is to avoid alcohol altogether — but that advice is explicitly about the risk to a developing fetus, not about the chance of a cycle working. So where does the near-universal clinic instruction come from? Largely from a different and entirely valid concern: the CDC states there is "no known safe amount of alcohol use during your pregnancy or when you are trying to get pregnant" — which is about preventing fetal alcohol spectrum disorder in a pregnancy nobody knows about yet. That is a fetal-safety argument, not a conception-probability argument. Both are legitimate. Conflating them is what produces guilt without producing babies.

Can you drink during stimulation or the two-week wait?

No study has measured this directly. The IVF cohorts all recorded intake before the cycle started; not one followed drinking through stimulation, retrieval or the luteal phase. So there is no evidence that a drink during stimulation changes an outcome, and no evidence that it does not.

What tips the answer is not fertility data. It is that from the moment of transfer — and arguably from retrieval, given how early implantation happens — a pregnancy may already exist and not yet be detectable. That is the situation the abstinence guidance is written for. The clean version of the advice is therefore: the stimulation-to-test period is the one stretch where stopping is straightforwardly justified, and it is justified on fetal-safety grounds rather than on any claim about your odds.

Does stopping actually change anything?

Here is the honest state of the reversibility evidence: it rests substantially on a single case report.

A man with alcohol-induced azoospermia — no sperm at all, with testicular biopsy showing maturation arrest — returned to "strictly normal semen parameters after no more than 3 months" following withdrawal (Sermondade et al., 2010). That is one patient, at an extreme starting point with maximal room to improve. It is also the most-cited piece of direct human evidence on this question, which tells you how thin the literature is.

And the intervention evidence stops short of outcomes. The one randomized trial here compared an intensive version of an mHealth coaching program against a light version, in 626 women undergoing IVF or ICSI plus 222 male partners. The intensive arm improved a composite nutrition-and-lifestyle score in which alcohol was one of five components; no alcohol-specific result was reported. Its endpoint was behaviour, not pregnancy, and the authors state that trials measuring live birth still need doing (Oostingh et al., 2020).

What the evidence does not show

And here is where the evidence runs out, which matters as much as where it holds:

• No randomized evidence exists that cutting alcohol before IVF improves live birth. Every causal claim in this space rests on observational data.
• No study has tested a cessation timeline against IVF outcomes. Any specific number is inference.
• Sperm DNA fragmentation is contested. One cross-sectional study of 211 samples found higher fragmentation in heavy drinkers than heavy smokers (Amor et al., 2022), while a meta-analysis of 40 studies and 23,258 men found no alcohol association with DNA fragmentation at all. Do not treat raised fragmentation from alcohol as established — the inputs with genuinely strong fragmentation evidence are covered in lowering sperm DNA fragmentation.
• Two of three well-powered studies found no independent binge-drinking effect.
• Alcohol is self-reported everywhere, and systematically under-reported — especially by women anticipating pregnancy, which biases results toward finding nothing.
• Alcohol travels with smoking, weight, diet and income. Nearly every author in this field acknowledges residual confounding.

What we do with this

A plan that matches the evidence rather than the anxiety:

1. For him: eliminate daily drinking and heavy sessions, starting about three months out. That is where the male signal actually lives — daily drinkers in the meta-analysis, above 25 units a week in the Danish cohort. Light weekly drinking has not been shown to matter. The three-month figure comes from the 64-day production line, not from an alcohol study.
2. For her: below roughly 4 to 6 US drinks a week, the fertility evidence does not show harm. Above that, it becomes suggestive but weak. The strongest reason to stop is not your conception odds — it is that an IVF cycle is undertaken in order to become pregnant, and there is no known safe amount in early pregnancy. That argument is clean and does not require any contested finding.
3. Set the stop date by the transfer, not by the anxiety. Stopping by the start of stimulation covers the period when a pregnancy becomes possible. Stopping six months early buys mental load, not biology.
4. If he flushes, treat him as a separate case and raise it with his physician.
5. Skip the heavy night around ovulation if you are also trying naturally between cycles — one cohort supports it and the cost is nothing.

If your cycle is closer than 3.5 months away, take the weeks you have. Some inputs matter even in the final weeks, and no window, used or unused, ever decided an outcome on its own.

Related reading

• The male fertility diet: what he eats in the 74 days before
• Lowering sperm DNA fragmentation in 74 days
• What to eat before IVF: a 90-day guide
• Is vaping safer than smoking before IVF?
• What nobody can tell you yet about preparing for IVF
• The research library — alcohol section

Frequently asked questions

How long before IVF should I stop drinking alcohol?

No study has tested a cessation timeline against IVF outcomes, so any specific number is inference rather than a finding. A defensible construction: roughly three months before the cycle for the male partner, because sperm take a mean of 64 days to develop and a change made later cannot fully appear in the sample; and by the start of stimulation for the female partner, because that is when a pregnancy becomes possible and there is no known safe amount of alcohol in early pregnancy.

Does alcohol reduce IVF success rates?

The evidence is genuinely inconsistent. A cohort of 2,545 couples found a live birth odds ratio of 0.84 for women drinking four or more drinks a week, but a Danish cohort of 1,708 women with complete registry follow-up found no association at any intake level. A 2022 dose-response meta-analysis of 26,922 women returned a maternal odds ratio of 0.83 with a confidence interval crossing the null. The paternal association with live birth was odds ratio 0.88, which did reach significance.

Does my husband need to stop drinking before IVF?

Cutting down is reasonable, but the case concerns heavy and daily drinking rather than any drinking. A meta-analysis of 40 studies and 23,258 men found no change in semen indices below about 7 units a week and harm above it. A separate meta-analysis of 16,395 men found the effect was driven by daily versus occasional drinkers, not never versus occasional. In natural conception, a pooled analysis of 2,679 couples found no association between male alcohol intake and fecundability at any level.

Can alcohol cause sperm DNA fragmentation?

This is contested rather than established. One cross-sectional study of 211 semen samples found higher DNA fragmentation in heavy drinkers, but a meta-analysis pooling 40 studies and 23,258 men found no association between alcohol and sperm DNA fragmentation at all. Smoking, by contrast, has a well-documented association with fragmentation. Present alcohol and sperm DNA as an open question rather than a known effect.

Is a glass of wine a week okay before IVF?

On the fertility evidence, no study has shown harm at that level. The largest natural conception cohort, 6,120 women, found no discernible effect below 14 servings a week. ASRM's concern begins above two drinks per day, and NICE tells men that 14 units a week spread across several days is unlikely to affect semen quality. The separate and stronger argument for stopping is that once a cycle begins, a pregnancy becomes possible, and there is no known safe amount of alcohol in early pregnancy.

Does binge drinking affect fertility more than regular drinking?

Two of the three well-powered studies that tested it found nothing. In 1,221 Danish men, binge drinking was not independently associated with semen quality. In 1,708 women undergoing fertility treatment, binge episodes were unrelated to live birth. One cohort of 413 women using daily diaries did find reduced fecundability per binge day in the luteal and ovulatory phases. The reasonable reading is to avoid heavy sessions around ovulation while recognising the evidence is thinner than commonly implied.

---

This article is educational and reflects associations reported in the scientific literature. It is not medical advice, not a diagnosis, and not a way to interpret your own results. As a Doctor of Acupuncture I offer lifestyle guidance that works alongside your physician and fertility clinic — I do not diagnose, interpret labs, or prescribe. Always make decisions about testing and treatment with your medical team. If you are concerned about your relationship with alcohol, your physician can connect you with the right support.