Ascertaining cases from a population-based tumor registry in which registration of primary lung cancer is virtually complete, we have consistently found that 40% to 50% of male lung cancer patients were farmers. We interviewed 273 newly diagnosed men and compared their occupational exposures, medical history and smoking characteristics to those of 187 male randomly selected community control subjects. We found that more of the control subjects were farmers (53.5% v 41.4%), that the control subjects tended to have larger farms (P less than .05), and that more control subjects spilled chemicals on their hands or clothing (47% v 28%, P less than .01) and had an accidental inhalation of a chemical directly into the lung (54% v 33%, P less than .005). We report an absence of correlation of lung cancer risk with occupational exposure to any specific pesticide or pesticides grouped by chemical composition. Adjusting for smoking pack-years or extent of pesticide use did not alter our preadjustment conclusions.