Background A possible connection between allergy and cancer has been suspected, but allergy-related conditions or atopy have been inconsistently associated with reduced risks of non-Hodgkin lymphoma. We investigated this association in a population-based case-control study and in a prospective study with prediagnostic blood specimens. Methods We carried out a population-based study of 3055 case patients with non-Hodgkin lymphoma and 3187 control subjects in Denmark and Sweden, including questionnaire information on allergy and blood specimens, and a nested case-control study within a prospective cohort of more than 400 000 Finnish women. In the second study, serum specimens from the 198 case patients who developed non-Hodgkin lymphoma within a median of 8.9 years after the blood was drawn were matched with serum specimens from 594 control subjects. In both studies, laboratory-based evidence of allergy (atopy) was determined in serum on the basis of specific IgE reactivity to common inhalant allergens. Dissemination of disease was classified by the Ann Arbor system. Odds ratios (ORs) with 95% confidence intervals (CIs) were calculated by logistic regression. Results In the first study, ever having hay fever, but not other allergic conditions, was associated with a reduced risk of non-Hodgkin lymphoma. In particular, subjects with specific IgE reactivity in serum had a 32% (95% CI = 20% to 42%) lower risk of overall non-Hodgkin lymphoma than those without such reactivity. However, among case patients, dissemination of the disease was strongly inversely associated with specific IgE reactivity. In the second (i.e., prospective) study, no association was found between non-Hodgkin lym-phoma and specific IgE reactivity, except possibly immediately before a diagnosis of non-Hodgkin lymphoma (≥10 years before diagnosis, OR = 1.00, 95% CI = 0.48 to 2.09; 5-9 years before, OR = 0.95, 95% CI = 0.50 to 1.84; 1-4 years before, OR = 0.33, 95% CI = 0.11 to 1.02; and <1 year before, OR = 0.27, 95% CI = 0.03 to 2.31). Conclusion Allergy may not be causally associated with the risk of non-Hodgkin lymphoma. The inverse association observed in some case-control studies may arise because non-Hodgkin lymphoma suppresses the immu-nologic response to allergens.