OBJECTIVES Assess direct (medical and drug) and indirect (workloss) costs of osteoporotic patients with non-vertebral (NV) fractures. METHODS Osteoporosis patients (ICD-9-CM code 733.0) were identified (1999-2006) from an employer claims database (~8,000,000 privately-insured beneficiaries; ages 18-64) and the Medicare Standard Analytic Files 5% sample (ages 65+). Osteoporotic patients with NV fracture (femur, pelvis, lower leg, upper arm, forearm, rib, or hip) were randomly matched on age, gender, employment status, and geography to osteoporotic controls. Annual direct costs were calculated (by third-party payer perspective) for privately-insured (n=4,764) and Medicare (n=48,742; medical costs only) beneficiaries. Indirect costs were calculated for privately-insured employees with disability data (n=1148). Costs were reported in 2006 dollars. RESULTS In Medicare, mean medical costs per non-vertebral, non-hip (NVNH) fracture patient were $7,463 in excess of controls ($13,720 vs. $6,258; p<0.01).The most expensive patients had fractures of the hip, multiple sites, and femur (excess costs of $25,120; $20,049; $19,213, respectively). Aggregate annual excess medical costs of these NVNH patients (n=35,933) were $268.1 million versus $200.9 million for hip fracture patients (n=7,997) (excludes patients with hip and NVNH on index date). In the privately-insured population, excess mean direct costs per NVNH fracture patient were $5,381 ($11,090 vs. $5,709; p<0.01). The most expensive patients had fractures of the hip, multiple sites, and pelvis (excess costs of $13,801; $9,642; and $8,164, respectively). Aggregate annual excess direct costs of these NVNH patients (n=4478) were $24.1 million versus $3.5 million for hip fracture patients (n=255). Mean excess indirect costs per NV employee were $1956 ($4,349 vs. $2393; p<0.01). CONCLUSIONS Excess direct and indirect costs of NV osteoporotic fracture patients to payers are substantial. While hip fracture patients are more costly per patient, NVNH fracture patients are associated with a larger percentage of fractures and aggregate excess costs for both these privately-insured and Medicare samples.