Experimental reproductive and developmental toxicity studies with polychlorinated biphenyls (PCBs) are reviewed in brief to determine their relevance for current environmental exposure of humans during the prenatal and postnatal developmental periods. Additional material is published in electronic form only, which contains graphic overviews on individual PCBs and various mixtures that are linked with the relevant citations. In this comprehensive article we focus on interactions of PCBs with biological substrates that could mediate adverse effects observed in experimental animals and in children, and the shortcomings of many of the animal studies available. A main point of criticism involves the relative lack of animal data on several of those persistent congeners, either as individual compounds or as environmentally relevant mixtures, which are currently used as a measure of human exposure. Experimental studies in animals are frequently conducted with commercial PCB mixtures, a test design that does not reflect the exposure situation in humans. Important improvements of animal experiments could be achieved by more complete reporting of litter data (pre- and post-natal losses, toxic signs in the dam and the offspring, birth weights and postnatal growth data), the inclusion of endpoints that have been found previously to be affected by PCBs, and measurements of internal exposure data.