The blood level of corticosterone was measured in mice following the injection on day 14 of pregnancy of a dose of corticosterone sufficient to cause a low frequency of cleft palate in the fetuses. This was compared with the blood levels present during maternal restraint and food deprivation that produced a similar frequency of cleft palate. The mean blood level over the 24 h following injection of corticosterone was 660 mug/100 ml, and during a similar period of restraint was 485 mug/100 ml. Other mice were subjected either to restraint or food deprivation for 24 h beginning day 14 of pregnancy, the plasma corticosterone levels measured during that time, and the frequency of cleft palate in late fetuses compared with the individual plasma corticosterone levels during treatment. There was a significant (P less than 0.025) correlation between high maternal corticosteroid levels and the frequency of cleft palate in the offspring of the restrained mice but not in the food-deprived animals. It is suggested that in some stressed mice endogenous plasma corticosterone can reach levels sufficient to account for the development of cleft palate.