There has been much concern recently about possible adverse health effects related to exposure to toxic chemicals among residents of Kanawha County in southern West Virginia. An epidemiological study of trends in cancer mortality from 1950-1984 among the general population of Kanawha County in southern West Virginia was mounted. Cabell County, West Virginia was chosen to be a comparison county for Kanawha in addition to West Virginia and the total United States. The cancer mortality rates for white males and females were calculated using NCHS mortality data and Census Bureau population data available on the Mortality and Population Data System (MPDS) at the University of Pittsburgh. Mortality rates for cancer in Kanawha and Cabell Counties were evaluated over the time period 1950-1984 with an age-period-cohort (APC) analysis. In this analysis, poisson regression models were fit using the statistical program GLIM (Generalized Linear Models) to determine the separate effects of age, period of death, and birth cohort on the specific cancers of interest (lung, liver, bladder, CNS, leukemia, lympho-reticulosarcoma, all cancers). There were no significant county differences for cancer death rates between Kanawha and Cabell Counties except for leukemia among white males [O.R. = 1.27, 95% (C.I. = 1.03-1.6)], and for lympho-reticulosarcoma [O.R. = 1.66(1.24-2.07)], suggesting a possible occupational exposure. For leukemia, aleukemia, the effect observed seems to have declined. In contrast, the elevation of lympho-reticulosarcoma rates has remained in recent years (1970-1984).