Many drugs with unknown safety profiles are administered to pregnant women, placing their offspring at risk. I assessed whether behavioral outcomes for children exposed during gestation to antidepressants, anxiolytics, anti-seizure, analgesic, anti-nausea and sedative medications can be predicted by more extensive animal studies than are part of the FDA approval process. Human plus rodent data were available for only 8 of 33 CNS-active drugs examined. Similar behavioral and cognitive deficits, including autism and ADHD emerged in human offspring and in animal models of these disorders after exposure to fluoxetine, valproic acid, carbamazepine, phenytoin, phenobarbital and acetaminophen. Rodent data helpful in identifying and predicting adverse effects of prenatal drug exposure in children were first generated many years after drugs were FDA-approved and administered to pregnant women. I recommend that enhanced behavioral testing of rodent offspring exposed to drugs prenatally should begin during preclinical drug evaluation and continue during Phase I clinical trials, with findings communicated to physicians and patients in drug labels.