Short answer: after repeated miscarriage, a short list of tests is worth having, a longer list is not, and the single most useful number is the one nobody leads with — for couples with unexplained recurrent miscarriage, the chance of a successful future pregnancy with supportive care alone is in the region of 75%.
That figure comes from the Royal College of Obstetricians and Gynaecologists, and it is the reason to read the rest of this page in a particular frame of mind. Recurrent miscarriage is devastating and it is usually not a sentence.
You will find both answers online, which is confusing, so here is exactly where each comes from.
RCOG — Green-top Guideline No. 17, fourth edition, June 2023 — defines recurrent miscarriage as three or more first-trimester miscarriages. But it adds an important clause: “clinicians are encouraged to use their clinical discretion to recommend extensive evaluation after two first trimester miscarriages, if there is a suspicion that the miscarriages are of pathological and not of sporadic nature.”
ESHRE — the European guideline, updated 2022 — sets the bar lower: “a diagnosis of Recurrent Pregnancy Loss (RPL) could be considered after the loss of two or more pregnancies.” It also counts very early losses that RCOG does not: a pregnancy confirmed by a positive blood or urine test qualifies, “including non-visualized pregnancy losses (biochemical pregnancy losses)”.
Two things both now agree on, which is a change from older advice:
ESHRE is candid enough to record its own internal disagreement: “some guideline group members would like to stress that they disagree with the suggested definition and will keep a definition of three or more consecutive losses in their clinical practice.”
Where we stand. We work to RCOG’s discretion clause rather than its headline number. If you have had two losses and there is any reason to think they were not simply bad luck — a second-trimester loss, a loss with a normal chromosome result, a known autoimmune or thyroid problem, your age — we will investigate rather than ask you to have a third.
Miscarriage itself is common and under-discussed. A 2021 Lancet analysis puts the pooled risk at 15.3% of all recognised pregnancies (95% CI 12.5–18.7).
Repeated loss is much less common:
Which is worth holding onto if you are in the last group: it is rare, and rare conditions get less attention than they deserve. It is also why RCOG notes that moving the definition from three losses to two “would more than double” the number of people classed as having recurrent miscarriage.
Drawn from RCOG GTG17 and ESHRE, with the grade each carries.
“Women with recurrent miscarriage should be offered testing for acquired thrombophilia, particularly for lupus anticoagulant and anticardiolipin antibodies, prior to pregnancy.”
This is the most important test on the list because it is the one with a treatment behind it. Antiphospholipid syndrome is treatable: “for women diagnosed with antiphospholipid syndrome, aspirin and heparin should be offered from a positive test until at least 34 weeks of gestation.”
“Women with recurrent miscarriage should be offered thyroid function tests and assessment for thyroid peroxidase (TPO) antibodies.” ESHRE agrees, and rates it a strong recommendation.
“Women with recurrent miscarriage should be offered assessment for congenital uterine anomalies, ideally with 3D ultrasound.” ESHRE calls 3D transvaginal ultrasound the preferred technique and says MRI is not the first-line option.
ESHRE added something in 2022 worth knowing about: “all women with RPL could have 2D ultrasound to rule out adenomyosis” — a condition of the womb muscle that is independently associated with miscarriage and is routinely missed.
This is the order most people get backwards. RCOG: “cytogenetic analysis should be offered on pregnancy tissue of the third and subsequent miscarriage(s) and in any second trimester miscarriage.”
Parental blood karyotyping comes after that, and only conditionally: offered “for couples in whom testing of pregnancy tissue reports an unbalanced structural chromosomal abnormality”, or where tissue testing was not possible.
So if you are being offered parental karyotyping as a first step, ask whether the pregnancy tissue can be tested instead. It answers a more useful question.
This is the section we think matters most, because these are the tests most often sold to people who are desperate.
“Inherited thrombophilias have a weak association with recurrent miscarriage. Routine testing for protein C, antithrombin deficiency and methylenetetrahydrofolate reductase mutation is not recommended.” ESHRE: do not screen for hereditary thrombophilia “unless in the context of research, or in women with additional risk factors”.
The exception is second-trimester loss, where Factor V Leiden, prothrombin gene mutation and protein S testing may be offered, “ideally within a research context”.
If you have been sold an MTHFR test for recurrent miscarriage, the guideline does not support it.
“Women with recurrent miscarriage should not be routinely offered immunological screening (such as HLA, cytokine and natural killer cell tests), infection screening or sperm DNA testing outside a research context.”
Natural killer cell testing in particular is widely marketed and sits outside guideline support. It is the kind of test that generates a worrying number and then a treatment to fix it, neither of which has been shown to help.
Named explicitly in the list above. ESHRE is slightly softer — it “could be considered for diagnostic purposes” — but the current UK fertility guideline, NICE NG257, goes further still for fertility generally: “Do not carry out testing for sperm DNA integrity (fragmentation).”
RCOG is unambiguous: “aspirin and/or heparin should not be given to women with unexplained recurrent miscarriage.” ESHRE agrees, with a strong recommendation, on the grounds that “there is no evidence that heparin alone, aspirin alone, or heparin in combination with low-dose aspirin improves the live birth rate in unexplained RPL”.
The distinction that matters: aspirin and heparin are the right treatment if you have antiphospholipid syndrome. They are not a general-purpose miscarriage treatment. That is exactly why the antibody test is the one worth having.
RCOG, citing a Cochrane review at the highest evidence level: various forms of immunotherapy “in women with unexplained recurrent miscarriage provides no significant beneficial effect over placebo in preventing further miscarriage… The use of immunotherapy should no longer be offered.”
ESHRE, separately and each rated strong: glucocorticoids “are not recommended”; there is “insufficient evidence to recommend intralipid therapy”; “no evidence to recommend Granulocyte colony-stimulating factor”; lymphocyte immunisation “should not be used… as it has no significant effect and there may be serious adverse effects”.
And NICE NG257, recommendation 1.42.1, published March 2026, applies the same conclusion to fertility treatment across the board: “Do not use immunological agents, including intralipids, intravenous immunoglobulins or steroids (glucocorticoids), as part of fertility treatment.” Its rationale: “there was no convincing evidence that immunological treatments are effective in improving fertility outcomes, and there are various safety concerns associated with their use.”
We do not offer any of these, and if you have been quoted for them elsewhere, bring the quote.
NICE guideline NG126 recommends it, and the conditions attached are specific:
“Offer vaginal micronised progesterone 400 mg twice daily to women with an intrauterine pregnancy confirmed by a scan, if they have vaginal bleeding and have previously had a miscarriage.” And: “if a fetal heartbeat is confirmed, continue progesterone until 16 completed weeks of pregnancy.”
Note the threshold is one previous miscarriage, not three. And note the two conditions: a scan-confirmed pregnancy, and bleeding.
That second condition is where the evidence sits. The PRISM trial, which produced this recommendation, found no significant overall benefit — live birth 75% with progesterone versus 72% with placebo, which did not reach statistical significance. The benefit appeared in the group with previous losses: among women with three or more previous miscarriages, live birth was 72% with progesterone against 57% with placebo.
NICE spells out the corollary: “there was evidence of no benefit in women with early pregnancy bleeding but no previous miscarriage, nor in women with previous miscarriage but no early pregnancy bleeding in the current pregnancy.”
So progesterone taken routinely from a positive test, with no bleeding, is not supported — which is worth knowing, because it is frequently prescribed that way. One piece of good news: since June 2025 this use of vaginal micronised progesterone is no longer off-label.
The part people actually want, and it is better than most expect.
RCOG, on unexplained recurrent miscarriage: “these women and their partners can be reassured that the prognosis for a successful future pregnancy with supportive care alone is in the region of 75%. However, the prognosis worsens with increasing maternal age and the number of previous miscarriages.”
Once a pregnancy is established, the odds shift sharply. In a classic cohort, if a fetal heartbeat had been identified by eight weeks, the chance of a successful outcome was 98%, rising to 99.4% at ten weeks. Getting past that first scan is genuinely the hurdle it feels like.
Over the longer run, a study of 987 women found that five years after their first consultation, 66.7% had achieved a live birth, rising to 71.1% after fifteen years. Age dominated: 81.3% of women aged 20 to 24 had a live birth within five years, against 41.7% of women aged 40 or over. Number of losses mattered too — 71.9% after three miscarriages, 50.2% after six or more.
An honest caveat ESHRE insists on: these prediction models “might overestimate the chances of a successful pregnancy”, and one widely used model performed poorly when tested on a new population. Which is why the right way to discuss your prognosis is against your own age and your complete pregnancy history, including live births and the order things happened in — not a figure from a webpage.
One number we cannot give you, because it does not exist in either guideline: a prognosis specific to exactly two unexplained losses. If someone quotes you one, ask where it came from.
We will take a full history — both partners, every pregnancy, in order — do the recommended tests and none of the unsupported ones, and tell you honestly when the answer is that no cause is found. “Unexplained” is the most common result and it carries the 75% figure above, which is not a consolation prize.
A female fertility assessment is £390, a couple assessment £495, and a saline scan of the womb cavity £300. Every price is on our prices page. Every result is reviewed by a Fertility Doctor before you are told anything.
Counselling is part of this, not an afterthought. Repeated loss is a bereavement that most people are expected to absorb quietly, often while being told it was “very early”. We have a fertility counsellor and you do not have to be in treatment to see her.
RCOG says three, with discretion to investigate after two where there is reason to suspect a cause rather than bad luck. ESHRE says two. We work to the discretion clause — if there is a reason, we will look.
Under ESHRE’s definition, yes, if the pregnancy was confirmed by a blood or urine test. Under RCOG’s, no — it counts first-trimester miscarriages. Either way, tell us about them; they are part of your history even where they do not change a definition.
No. RCOG advises against it outside research, and there is no treatment based on the result that has been shown to work. We would rather tell you that than charge you for it.
If you have a scan-confirmed pregnancy and you bleed, yes — 400 mg twice daily until 16 weeks if a heartbeat is seen, which is exactly what NICE recommends. Without bleeding, the evidence does not support it.
No. Not lifting, not exercising, not stress, not working, not the glass of wine before you knew. Where a cause is found it is usually antiphospholipid syndrome, a chromosomal issue in the pregnancy, a thyroid problem or the structure of the womb — none of which anybody causes.
It changes the urgency. The prognosis figures above fall steeply with age, so if you are over 35 with repeated losses, investigation should not wait for another one.
Not in itself, and be careful with clinics that imply it can. NG257 recommends against genetic testing of embryos to improve live birth rates. IVF has a role where there is a separate fertility problem, or a known chromosomal rearrangement in one partner — not as a general treatment for recurrent loss.
If you have had repeated losses and nobody has taken a proper history, start with a free 15-minute consultation with a Fertility Specialist — no referral needed, no obligation. Book a free consultation, or read about our fertility assessments.
