Emerging data suggest that inflammation may play a role in the etiology of diabetes mellitus. Because few prospective studies have addressed this issue, the author examined the relation between leukocyte count and erythrocyte sedimentation rate and diabetes incidence using data from the National Health and Nutrition Examination Survey Epidemiologic Follow-up Study (from 1971-1975 to 1992-1993). Of 8,352 participants included in the analysis, 878 developed incident diabetes during the approximately 20-year follow-up. After adjustment for age, smoking status, systolic blood pressure, cholesterol concentration, use of antihypertensive medication, recreational exercise, non-recreational activity, alcohol use, and body mass index, the hazard ratios from proportional hazards for participants with a leukocyte count of > or = 9.1 x 10(9)/liter compared with participants with a leukocyte count of < or = 5.7 x 10(9)/liter were 1.33 (95% confidence interval (CI): 0.81, 2.19) for men and 1.68 (95% CI: 1.21, 2.34) for women. The adjusted hazard ratios for participants with an erythrocyte sedimentation rate of > or = 26 mm/hour compared with participants with an erythrocyte sedimentation rate of < or = 5 mm/hour were 1.85 (95% CI: 0.97, 3.54) for men and 0.83 (95% CI: 0.47, 1.44) for women. These results provide limited support to the hypothesis that inflammation is an etiologic factor for diabetes.