This analysis represents a 5-year update of our mortality study of 7075 PCB exposed capacitor workers that now includes 1654 deaths and 235,984 person-years of observation with follow-up through 1998. In hourly males and females the observed number of deaths from all-cancers and all-causes were similar to the expected numbers. In salaried males all-cause and all-cancer mortality were significantly below the expected. In salaried females, all-cause mortality was significantly below the expected and all-cancer mortality was below the expected, but not significantly. We again failed to find any significant excess mortality in the a priori cancers of concern or in any other cancers in the total cohort or in the highly exposed portion of the cohort. These results expand on our previous observations and as before the data fail to demonstrate any causal association between occupational PCB exposure and excess cancer mortality or any other causes of death.