The mortality experience of agricultural extension agents in the Cooperative Extension Service (CES) of the U.S. Department of Agriculture who died during the period January 1, 1970-December 31, 1979 (n = 1,495 white males) was evaluated in proportionate-mortality and case-control studies. The proportionate-mortality analysis was used to identify cancers that might be elevated in this occupational group compared with the U.S. white male population. All cancers with a significantly elevated proportionate-mortality ratio were more thoroughly evaluated in the case-control study, where there is presumably less of a selection bias in the comparison. In the case-control study, leukemia demonstrated a statistically significant linear trend with duration of employment as an extension agent. Smaller, but nonsignificant, trends were seen for non-Hodgkin's lymphoma, multiple myeloma, and brain cancer. The odds ratio for Hodgkin's disease and cancers of the colon, prostate, and kidney did not vary with the number of years on the job. These patterns resemble cancer risks seen among farmers, suggesting that agricultural factors may also play a role in the origin of these tumors among extension agents.