Most fractures occur in postmenopausal women who do not have osteoporosis by bone density criteria (T-score>-2.5). Prevalent vertebral deformity is a strong risk factor for incident fractures independent of bone mineral density, yet the majority of these deformities are clinically unapparent. Spine imaging on a dual-energy densitometer, called vertebral fracture assessment (VFA), can accurately detect these deformities. The purpose of this modeling study was to estimate the lifetime societal costs and health benefits of VFA and confirmatory follow-up radiography to detect prevalent vertebral deformity in osteopenic (T-score -1.5, -2.0, or -2.4) postmenopausal women, followed by anti-resorptive drug therapy for those with one or more deformities. We compared three strategies; no initial drug therapy, 5 yr of initial alendronate therapy, or VFA followed by 5 yr of alendronate therapy in those with one or more vertebral deformities confirmed on radiography (VFA strategy). Results for the base-case analyses showed that the cost per quality adjusted life year (QALY) gained for the VFA strategy relative to no initial drug therapy ranged from 18,000 US dollars (for a 60-yr-old with a femoral neck T-score of -2.4) to 77,000 US dollars (for an 80-yr-old with a T-score of -1.5). VFA with selective confirmatory radiography is cost-effective, assuming a societal willingness to pay per QALY gained of 50,000 US dollars, for postmenopausal women aged 60 to 80 yr with femoral neck T-scores between -2.0 and -2.4, and for women age 60 or 70 yr with a T-score of -1.5. Assuming a societal willingness to pay of 100,000 US dollars per QALY gained, VFA is also cost-effective for women age 80 yr with a T-score of -1.5. These conclusions are robust to reasonable changes in fracture rates, assumed fracture disutility, discount rates, relative risk of fracture attributable to vertebral deformity, alendronate cost, and drug adherence.