The objective of this paper is to determine if phenobarbital exposure during pregnancy affects developmental outcome at age 2 years. Between 1991 and 1994, 401 pregnant patients at risk for delivery prior to 34 weeks' gestation were invited to participate; 48 mothers declined entry. Before delivery, pharmacy randomized the pregnant women to receive phenobarbital and vitamin K or identically appearing placebo in a blinded fashion. Developmental follow-up at age 2 years was performed. Children from the treatment group scored significantly lower on the Bayley Mental Developmental Index (mean MDI +/- 1 SD) than children whose mothers were randomized to the placebo group [104 +/- 21 (n = 59) vs. 113 +/- 22 (n = 62), p = 0.023]. Of 36 independent variables, randomization group was one of five that individually contributed to the prediction of the Bayley MDI score (p < 0.05). It was concluded that perinatal phenobarbital therapy may impair developmental outcome.