OBJECTIVES: To determine the cross-sectional and longitudinal associations between 25-hydroxyvitamin D (25(OH)D) levels and frailty status in older men. DESIGN: Prospective cohort study. SETTING: Six U.S. community-based centers. PARTICIPANTS: One thousand six hundred six men aged 65 and older. MEASUREMENTS: 25(OH)D (liquid chromatography tandem mass spectroscopy) and frailty status (criteria similar to those used in the Cardiovascular Health Study) measured at baseline; frailty status assessment repeated an average of 4.6 years later. Frailty status was classified as robust, intermediate, or frail at baseline and robust, intermediate, frail, or dead at follow-up. RESULTS: After adjusting for multiple potential confounders, men with 25(OH)D levels less than 20.0 ng/mL had 1.5 times higher odds (multivariate odds ratio (MOR)=1.47, 95% confidence interval (CI)=1.07-2.02) of greater frailty status at baseline than men with 25(OH)D levels of 30.0 ng/mL or greater (referent group), whereas frailty status was similar in men with 25(OH)D levels from 20.0 to 29.9 ng/mL and those with levels of 30.0 ng/mL or greater (MOR=1.02, 95% CI=0.78-1.32). However, in 1,267 men not classified as frail at baseline, there was no association between lower baseline 25(OH)D level and odds of greater frailty status at the 4.6-year follow-up. Findings were the same when 25(OH)D was expressed in quartiles or as a continuous variable. CONCLUSION: Lower levels of 25(OH)D (<20.0 ng/mL) in community-dwelling older men were independently associated with greater evidence of frailty at baseline but did not predict greater risk of greater frailty status at 4.6 years.