Frequently occurring nonepileptic psychogenic seizures (PNES) are a cause of substantial morbidity. Differentiation of these from true seizures may sometimes be very difficult. Serum prolactin level estimation following the event has been described as a useful test for this purpose. We conducted this study to assess the role of this test in diagnosis of PNES. Serum prolactin was estimated from venous blood samples of 19 patients (13 females, 6 males) with PNES and 17 patients (5 females, 12 males) with true complex partial seizures with or without secondary generalization. The age range was 12-39 years in the PNES group and 9-42 years in the true seizure group. Five patients (all females) in the PNES group (26.3%) had raised prolactin levels, all of them having greater than twice normal levels. In the true seizure group, 10 of 17 (58.8%) patients had raised levels; only 3 (17.6%) of these had greater than twice normal levels. The difference in percentage of patients with abnormal prolactin levels between these groups was not found to be significant. We demonstrate that serum prolactin level estimation is not a useful method for differentiation of psychogenic nonepileptic from true epileptic seizures.