There are limited data in the literature concerning risk factors for incident fractures in men with rheumatoid arthritis (RA). We evaluated the association between potential risk factors and incident clinical fractures in male Japanese patients with RA. A total of 1050 male patients with RA were enrolled in a prospective, observational cohort study from 2000 to 2005. Participants were followed from 6 to 66 months (median follow-up, 48.7 months) and classified into three groups according to their incident fracture status from baseline: no new fracture, any new nonvertebral fracture, and new clinical vertebral fracture. The associations of potential risk factors were analyzed by Cox proportional hazards models. During follow-up, 30 patients (2.9%) developed a new nonvertebral fracture or a vertebral fracture. The baseline age, history of total knee replacement (TKR), and serum C-reactive protein (CRP) levels were associated with any nonvertebral fracture [baseline age: hazard ratio (HR), 1.08, 95% confidence interval (CI), 1.03-1.14; history of TKR: HR 6.02, 95% CI 1.19-30.42; and CRP: HR 0.60, 95% CI 0.38-0.95]. The baseline Japanese health assessment questionnaire (HAQ) score and daily dose of prednisolone were also associated with the incidence of clinical vertebral fractures (HR 7.74, 95% CI 2.10-28.48, and HR 1.28, 95% CI 1.14-1.45, respectively). Older age, history of TKR, and low serum CRP levels appear to be associated with any incident nonvertebral fracture in Japanese men with RA. High HAQ disability score and baseline doses of daily prednisolone may correlate with incident clinical vertebral fracture in Japanese men with RA.