All occupational cancer epidemiology articles that reported data on non-whites from four major journals were identified for the years 1984-1987. In addition to these 14 papers, four more papers were identified from computer searches, and another 13 were found by following up references from the original 14. Within these 31 papers, there were 36 analyzable cohorts (subgroups by job and sex). Five cohorts had elevated all-cancer mortality ratios in non-whites without an increase in whites. Sixteen cohorts showed elevated ratios in both racial groups, with 11 of the 16 having a higher non-white cancer mortality ratio than white cancer mortality ratio. For the 25 studies in which at least one racial group's cancer ratio was elevated, the non-white ratio exceeded the white ratio in 17 (68%). The overall paucity of articles on non-white cancer mortality, as well as the apparent trend toward increased occupational cancer mortality in non-whites, is discussed.