A prospective case-control study was performed to determine whether inhibin A concentration is a clinically useful marker of ectopic pregnancy (EP). Inhibin A concentration in patients diagnosed with EP by laparoscopic and pathological findings (n = 17) was compared with that in missed miscarriage (n = 35), incomplete miscarriage (n = 14), spontaneous miscarriage (n = 5), threatened miscarriage (n = 6), normal pregnancy (n = 24) and non-pregnant controls (n = 20). The data were analysed using the Mann-Whitney U-test. EP yielded significantly lower inhibin A concentrations compared with normal pregnancy, 12.7 +/- 11.7 versus 237.3 +/- 125.9 pg/ml (P < 0.0002), and similar concentrations to non-pregnant controls (13.3 +/- 14.3 pg/ml). Inhibin A concentrations in abnormal pregnancies were significantly lower than in the normal pregnancy group: missed miscarriage 42.4 +/- 54.9 pg/ml (P < 0.0002); spontaneous miscarriage 47.5 +/- 55.6 pg/ml (P < 0.0002); and incomplete miscarriage 12.2 +/- 10.5 pg/ml (P < 0.0002). Threatened miscarriage was not statistically different to normal pregnancy (183.1 +/- 119.4 pg/ml). Human chorionic gonadotrophin concentrations in EP were not statistically significantly different compared with missed miscarriage and incomplete miscarriage. In conclusion, serum inhibin A concentration may be a reliable marker of EP.