A case/control study was carried out of cancer associated mortality at a transformer assembly factory in Massachusetts. Seven types of exposure with mutagenic or carcinogenic potential were rated together with the operations in which they were used. The chemicals were pyranol (1336363), benzene (71432), trichloroethylene (79016) (TCE), mixed solvents, machining fluids, asbestos (1332214), and synthetic resins. The cohort consisted of 1,821 white male subjects with 51,063 person years of employment. Site specific cancer deaths were regarded as cases, and comparisons were selected from deaths considered unassociated with any of the exposures. Odds ratios (ORs) were calculated. Results showed that 28 exposure/cancer associations could be identified as important in the binary exposure regressions or in the contingency table screening. Pyranol lymphomas and solvent oral/laryngeal/pharyngeal cancers showed a greatly reduced association, whereas a consistent positive association was evident between benzene and brain tumors, other solvents and lymphomas, and resins with lung cancer. Regressions with multiple exposures were also computed. A modification of the induction latency analysis method was used, and the only notable finding was an association between machining fluids and kidney cancer. Associations of resins with rectal and lung cancer, machining fluids with kidney cancer, and TCE with leukemias seemed to be largely or entirely concentrated to post 1950 exposures. Cross tabulation showed a similar association between level 2 solvents and reticulosarcoma. The authors conclude that the link between machining fluids and kidney cancer is new, and that this as well as a possible association between lung cancer and resin systems deserve further investigation.