Recent epidemiological studies have suggested an increased risk of non-Hodgkin lymphoma (NHL) from carbamate insecticide use among farmers. To further explore the possible relationships, we conducted a pooled analysis of three population-based case-control studies conducted in four midwestern states in the United States. A total of 985 white male subjects and 2895 control subjects were included in this analysis. Unconditional logistic regression was used to estimate the association and control for confounding. Compared with nonfarmers, farmers who had ever used carbamate pesticides had a 30% to 50% increased risk of NHL, whereas farmers without carbamate pesticide use showed no increased risk. Analyses for individual carbamate pesticides found a more consistent association with Sevin but not carbofuran, butylate, or S-ethyl dipropylthiocarbamate plus protectant. Among farmers using Sevin, the risk of NHL was limited to those who personally handled the product, those who first used the product for > or = 20 years before their disease diagnosis, and those who used the product for a longer period. These associations persisted after adjusting for other major classes of pesticides. These results suggest an increased risk of NHL associated with carbamate pesticide use, particularly Sevin. Further investigation of the association is warranted.