A study of 1,497 nonmetropolitan counties was conducted to test the possible contribution of agricultural chemical use to cancer mortality rates in rural counties. The dependent variables were 20-year age-adjusted mortality rates for 1950 to 1969 for five categories of cancer: genital, urinary, lymphatic, respiratory, and digestive. Because sex and race are both recognized factors in cancer etiology, only rates for white males were used in the analysis. Significant associations were found between agricultural chemical use and four of the five types of cancer examined. Herbicides were positively and significantly associated with genital, lymphatic, and digestive cancer. Insecticides had a strong positive relationship to respiratory cancer. Fertilizer use was largely unrelated to cancer mortality with the exception of a modest negative association with digestive cancer. Of the remaining social, economic, and demographic variables, foreign stock exhibited the largest and most consistent influence and was the strongest predictor of digestive cancer. Income and education were both significant in respiratory and digestive cancer, but of opposite sign. In spite of the limitations of this study, the findings are highly suggestive of the need for additional research on possible links between agricultural chemical use and county cancer mortality. (JHZ)