The hypothesis that in utero exposure to diazepam increases the risk of oral-cleft anomalies was evaluated in a case-control study, in which 445 infants with cleft lip with or without cleft palate and 166 with cleft palate without cleft lip (cleft palate alone) were compared with 2498 control infants having other birth defects. For exposure to diazepam during lunar months 1 through 4 relative to no exposure during pregnancy, the estimated relative risks were 1.0 for cleft lip with or without cleft palate (95 per cent confidence interval, 0.5 to 2.1) and 0.8 for cleft palate alone (0.3 to 2.7). After control for all identified potential confounding factors, the corresponding estimates were 0.8 (0.4 to 1.7) and 0.8 (0.2 to 2.5), respectively. The findings were unchanged when maternal suspicion that diazepam might be a teratogen was taken into account. The data suggest that first-trimester exposure to diazepam does not materially affect the risk of cleft lip with or without cleft palate or of cleft palate alone.