OBJECTIVE -To examine the epidemiologic association between diabetes and risk of nonHodgkin's lymphoma (NHL).RESEARCH DESIGN AND METHODS -We searched MEDLINE for observational studies on the association between diabetes and NHL in adults using the keywords "diabetes" and "lymphoma." Prospective cohort studies that reported relative risks or standardized incidence ratios and case-control studies that reported odds ratios with 95% CIs were included. A randomeffects model was used to combine results from the individual studies.RESULTS -A total of 15 manuscripts (reporting data from 5 prospective cohort and 11 case-control studies) met the inclusion criteria. Combining data from all studies, the risk ratio (RR) of developing NHL in patients with diabetes was 1.19 (95% CI 1.04 -1.35). Based on prospective studies, patients with diabetes had an RR of developing NHL of 1.41 (1.07-1.88), without heterogeneity among studies (I 2 ϭ 34.3%; P Ͼ 0.10). Based on case-control studies, patients with diabetes had an RR of 1.12 (95% CI 0.95-1.31) of developing NHL compared with people without diabetes, with some heterogeneity among studies (I 2 ϭ 36.28%; P ϭ 0.09).CONCLUSIONS -Diabetes is associated with a moderately increased risk of NHL, which is consistent with other reported associations between diabetes and malignancies. Future studies should focus on elucidating potential pathophysiologic links between diabetes and NHL.