We conducted a randomized trial comparing expectant management versus immunotherapy with paternal leukocytes to improve obstetric outcome in women with unexplained recurrent abortion. Eligible for the study were women with unexplained recurrent abortion (three or more miscarriages and no live birth), negative findings of immunological screening and no inhibition of the mixed lymphocyte culture. These women were seen for the first time between October 1988 and March 1991 in a network of obstetric departments in Northern Italy. Subjects positive for HLA DR3 or with a partner positive for hepatitis virus B antigen were not eligible. A total of 44 women entered the study. Patients were randomly allocated to immunotherapy (22 women) or expectant management (22 women). Women allocated to immunotherapy were given 200 x 10(6) purified paternal lymphocytes before pregnancy. Median follow-up was 24 months (range 10-39) in the immunotherapy group and 25 months (range 11-38) in the expectant management group. Out of the 22 women randomized to immunotherapy, 16 became pregnant and the corresponding value was 14 in the expectant management group. Spontaneous abortion occurred in six out of the 16 pregnancies observed in the treated women. Among the 14 pregnancies observed in the expectant management group, two aborted and one late fetal death occurred. The cumulative proportions of women who became pregnant over 4 years were 37 and 45% in the immunotherapy and expectant management groups respectively; this difference was not significant. No adverse effect was observed in treated women.