A population-based case-control study was conducted between 1988 and 1995 in the San Francisco Bay Area of California to determine risk factors for non-Hodgkin's lymphoma. Participants completed in-person interviews, and Wood was drawn to test for viruses and lymphocyte subsets. This report includes data for 1, 281 cases and 2, 095 controls. In multivariate analyses, the factors associated with a decreased risk for non-Hodgkin's lymphoma were allergy to plants, bee and wasp stings, five or more vaccinations, drugs to lower blood cholesterol, nonsteroidal anti-inflammatory drugs, total number of sexual partners, and lifetime marijuana use, whereas an increased risk was associated with cimetidine and other histamine H2-receptor antagonists, splenectomy, gonorrhea, and body mass index. Unique to sex-specific models was an increased risk for endocrine gland disorders among women and for polio among men. Median CD3, CD4, CD8, CD20, and lymphocyte counts for non-Hodgkin's lymphoma patients were significantly lower than those for controls. These results implicate environmental factors that may influence the early stages of lymphomagenesis by stimulating the immune system. Antigen-driven B cells that accumulate to form lymphoma may be suppressed by immunologic stresses such as exposure to an increased number of sexual partners and to certain medications. A history of allergies provides evidence for a persistent capacity for B-cell differentiation and therefore a decreased accumulation of B cells. The decreased risk for non-Hodgkin's lymphoma with use of nonsteroidal anti-inflammatory drugs and cholesterol-lowering drugs is consistent with a macrophage inflammatory role in B-cell proliferation. Am J Epidemiol 1999; 150: 375–89.