The aim of the present study was to evaluate whether serum activin A levels may represent, in addition to intact human chorionic gonadotrophin, a marker of placental tumors in the course of chemotherapy. Serial determinations of serum levels of activin A were performed in women with hydatidiform mole (n = 2) or choriocarcinoma (n = 3). Serum activin A levels were measured by using a new specific two-site enzyme immunoassay (EIA) able to detect the dimeric, bioactive, form of the protein. Serum hCG concentrations in samples taken after evacuation before starting chemotherapy were greater than in healthy non-pregnant women (p < 0.001) and decreased following chemotherapy. Activin A serum levels in women with trophoblastic disease after evacuation were significantly higher than in healthy non-pregnant women, but chemotherapy did not significantly affect circulating levels. No correlation was found between changes of activin A and total hCG serum concentrations. Measurement of activin A by ELISA in presence of persistent molar tumor does not seem to be of clinical interest in the follow-up of disease, resulting activin A concentrations after chemotherapy in the range of values occurring throughout menstrual cycle. These evidences suggest that hCG determination is still the most valid for follow-up, because only intact hCG could detect the persistence of trophoblast tissue.