PURPOSE: There is limited data pertaining to osteoporotic fractures among North American women of South Asian (SA) descent. This study examines fracture incidence and risk factors among post-menopausal SA, Chinese and White women undergoing mineral density (BMD) testing within a large healthcare organization in Northern California. METHODS: Using data from a retrospective study of women aged 50-85 years with femoral neck BMD measured between 1997 and 2003, we identified a subset of women of SA race and an age-matched subgroup of Chinese (1:5) and White (1:10) women and examined rates of incident wrist, humerus and hip fractures up to 10 years following BMD. Clinical and demographic risk factors were identified using health plan databases. Multivariable Cox regression analyses were conducted to examine predictors of incident fractures. RESULTS: The study cohort included 449 SA, 2245 Chinese and 4490 White women, with an average age of 58.4 +/- 6.1 years. The prevalence of femoral neck osteoporosis was higher among SA (8.9%) compared to White (6.5%) women and tended to be lower than Chinese (11.9%) women. More SA (7.1%) and White (9.6%) women had prior fracture compared to Chinese women (4.5%) and racial differences in smoking, rheumatoid arthritis, glucocorticoid use and hormone replacement therapy were seen. During a median of 8.4 years follow-up, wrist fracture incidence was similar among SA and White women (286 and 303 per 100,000 person-years, respectively) but significantly lower among Chinese women (130 per 100,000 person-years). In multivariable analyses, lower BMD, prior fracture and White and SA race (compared to Chinese race), were associated with a higher relative rate of wrist fracture. Lower BMD, prior fracture, older age and White but not SA race were also associated with a higher relative rate of non-vertebral (wrist, humerus or hip) fractures. CONCLUSIONS: Post-menopausal South Asian women differed from Chinese and White women with respect to prevalence of femoral neck osteoporosis, certain risk factors and site of osteoporotic fracture. These findings support the need for more studies examining fracture risk and outcomes specific to SA women residing in the U.S. to inform clinical decisions relevant to fracture risk.