In an attempt to further develop basic principles to guide research in neurobehavioral teratology, six experiments were conducted to examine the effects of prenatal haloperidol (a D2 dopamine antagonist) exposure on striatal D1 and D2 binding sites. Another laboratory has repeatedly reported that prenatal exposure to this dopamine antagonist reduces striatal dopamine binding sites in exposed offspring. Our initial studies were successful in replicating and extending these previously reported reductions in D2 dopamine binding sites in caudate of rats exposed prenatally to haloperidol. However, additional experiments in our laboratory, in which pups were exposed to a range of haloperidol doses over gestational periods when the dopamine system has been reported to be most vulnerable to prenatal haloperidol exposure effects, have repeatedly failed to replicate our initial findings. Three other laboratories have also failed to duplicate this effect. The results of these studies suggest that beyond "standard" confounding variables, neurobehavioral teratologists are faced with as yet poorly understood factors that influence replication of findings within and between laboratories. These findings also emphasize the importance of within- and between-laboratory replication of experimental findings.