The current policy for regulating polychlorinated biphenyls (PCBs) is based on one chronic bioassay that examined the carcinogenicity of a 60% chlorinated PCB (Norback & Weltman, 1985). All studies originally considered by the U.S. Environmental Protection Agency (EPA) for use in calculating a cancer slope factor (CSF) for PCBs were reevaluated and new CSFs calculated based on the results of a pathology reassessment (Moore et ah, 1994). When studies of 60% chlorine PCBs from 3 different laboratories were compared, there was no scientific basis for selecting only 1 data set for deriving CSF estimates. Using a geometric mean to calculate a CSF based on all studies of PCBs with 60% chlorine replaces the current value of 7.7 (mg/kg/d)−1with a value of 1.9 (mg/kg/d)−1. CSFs for PCBs containing less than 60% chlorine (54% and 42%) were less than 1.0 (mglkgld)−1. Using a toxic equivalency factor (TEF) approach similar to that of 2,3,7,8‐tetrachlorodibenzo‐p‐dioxin shows no correlation between toxic equivalent dose and CSFs, indicating that use of TEFs is not predictive of cancer potency for PCBs. Based on these findings, PCB cancer risk assessment policy would more closely reflect scientific data if (1) separate risk assessments were developed for each major PCB formulation and (2) all appropriate data were used when calculating cancer potency for PCB mixtures of 60% chlorine.