The relation between reproductive pattern and the risk of gestational trophoblastic disease was evaluated in a case-control study conducted in Northern Italy on 310 women with histologically confirmed gestational trophoblastic disease and two control groups consisting of 290 obstetric subjects and 394 patients in hospital for acute, nonobstetric, nongynecologic conditions. Compared to that for nulliparous women, the estimated age-adjusted relative risk of trophoblastic disease for parous women was 0.6 (90% confidence limit = 0.4 to 0.9) when obstetric controls were used as a comparison group and 0.4 (95% confidence limit = 0.2 to 0.6) compared with other controls. Conversely, a history of spontaneous abortions was associated with elevated risk of gestational trophoblastic disease, and the risk increased significantly with increasing number of spontaneous abortions. When the combined effect of parity and spontaneous abortions was considered, the major factor influencing the risk of gestational trophoblastic disease was the existence of one or more previous term pregnancies.