A case control study, based on death certificates, was conducted of non Hodgkin's lymphoma (NHL) and Hodgkin's disease (HD) among white males in Hancock County, Ohio. Cases were defined as decedents for whom NHL or HD was coded as cause of death, identified among all white male residents, aged 15 or over, who died between 1958 and 1983. Four comparisons per case were chosen as a random sample stratified by age at death and year of death. Cases and comparisons were compared with respect to usual occupation and industry, and summary odds ratio (OR) and chi square significance tests were calculated. The study involved 61 cases of NHL, 15 cases of HD and 304 comparisons. Farming was the largest single occupation and agriculture the largest industry among the study subjects. Farmers constituted 25, 20 and 16 percent of NHL cases, HD cases and comparisons, respectively. The overall summary odds ratios (ORs) for NHL and HD among farmers were 1.6 for NHL and and 2.7 for HD. Thirteen of the 15 farmers with NHD died between 1958 and 1973, giving an OR of 2.1 for this time period. The OR for HD in the year of death stratum 1958 to 1963 for farmers aged 15 to 64 was 21.2, with all three observed cases occurring in this time period. The authors conclude that the number of excess deaths from NHL among farmers observed in this study is insufficient to explain the high rate of NHL seen in Hancock County from 1960 to 1979.