Polychlorinated biphenyls (PCBs) are a mixture of chemicals. Some congeners of the mixture are highly persistent both in the environment and in humans. Although PCBs have not been used commercially since about 1977 in the US, they can still be detected in human blood and tissues in this country. PCB levels are declining and are often no longer detectable in younger people. A cursory review of recent animal studies is provided. Studies to determine whether PCBs cause cancer in humans, neurobehavioural effects, abnormal thyroid and immune function in children and low birth weight are discussed in more detail. These studies are inconclusive and do not provide clinical evidence that PCBs at levels encountered with human exposure produce adverse health effects. The differences in PCB blood or tissue concentrations between controls and cases, or between the upper and lower end of various environmentally exposed groups of children or adults, are small. Although some effects are statistically significantly different, they do not appear to be biologically significant. Many studies on the effects of PCBs are difficult to interpret because the range of normal values for clinical and neurobehavioural tests are not provided or appropriately considered, there was no, or inadequate, control for potential confounders. In occupational mortality studies, exposures were much higher. In some studies, various specific cancers were elevated. However, these appear to be chance observations resulting from multiple comparisons since the increase of specific cancers was not consistent between studies and was no longer present in some cohorts when studies were repeated at a later date with longer follow-up. Overall, the data fail to demonstrate conclusive adverse health effects of PCBs at concentrations encountered with human exposures.