BACKGROUND: All published studies of health care utilization in gout have been cross-sectional to date, and most used a patient-reported diagnosis of gout. Our objective was to assess health care utilization and its predictors in patients with physician-confirmed gout in a prospective cohort study. METHODS: In a multi-center prospective cohort study of U.S. veterans with rheumatologist-confirmed gout (N = 186; two centers), we assessed patient self-reported overall and gout-specific health care utilization with the Gout Assessment Questionnaire (GAQ) every 3-months for a 9-month period. Comparisons were made using the student's t test or the chi-square, Wilcoxon rank sum test or Fisher exact test, as appropriate. Mixed effects Poisson regression was used to assess potential correlates of gout-related health care utilization. RESULTS: Mean age was 64.6 years, 98% were men, 13% Hispanic or Latino, 32% were African-American, 6% did not graduate high school, mean serum urate was 8.3 and mean Deyo-Charlson score was 3.1. During the past year, mean gout-related visits were as follows: rheumatologist, 1.5; primary care physician, 2 visits; ≥1 inpatient visits, 7%; ≥1 ER visits, 26%; and urgent care/walk-in visit, 33%. In longitudinal analyses, African-American race and gout flares in the last 3 months were associated with significantly higher rate ratio of gout-related outpatient visits. African-American race and lack of college education were associated with significantly higher rate ratio for gout-related urgent visits and overnight stays. CONCLUSIONS: African-American race and recent gout flares were associated with higher outpatient utilization and African-American race and no college education with higher urgent or inpatient utilization. Future studies should examine whether modifiable predictors of utilization can be targeted to reduce healthcare utilization in patients with gout.