The possibility that antiemetics used during pregnancy are teratogenic has been hotly debated; the effect of vomiting itself, however, has been largely ignored. The relation between vomiting and congenital malformations was examined in a prospective study of 16398 women who registered for prenatal care at or before 20 weeks' gestation. The odds ratios for malformations among women who vomited compared with women who did not were 1.14 for major malformations (p = 0.13), 0.88 for deformations (p = 0.39), 1.03 for hernias or undescended testes (p = 0.82), 1.06 for any of these three conditions (p = 0.38), 1.09 for minor anomalies (p = 0.14), and 1.10 for any anomaly (p = 0.03). After adjustment for use of antiemetics and five other confounding variables vomiting was not associated with a significantly increased risk of any of the above malformations. These data suggest that the increased risk, if any, among women receiving antiemetics during pregnancy is due to the drugs, not the vomiting.