Background: People living with HIV (PLWH) are at higher risk of reduced bone mineral density (BMD) and fragility fracture compared to the general population. Possible causes include: an increased prevalence of general fracture risk factors (GFRFs) in PLWH; a direct effect of HIV; and a contributory role of antiretroviral therapy. HIV guidelines recommend FRAX® (www.sheffield.ac.uk/FRAX) for fracture risk assessment in PLWH. FRAX®, however, incorporates GFRFs but not HIV disease-specific factors. Hypotheses: 1. Both HIV disease-specific factors and GFRFs contribute to reduced BMD and fracture risk in PLWH. 2. FRAX® correlates poorly with BMD in PLWH. Methods: Phase One: The prevalence of GFRFs and fractures were recorded in PLWH. FRAX® 10-year osteoporotic fracture probabilities (FRAX® scores) were calculated. Phase Two: A subset of the Phase One cohort were recruited proportionately by race, gender and FRAX® scores (low, intermediate and high) for dual-energy X-ray absorptiometry BMD measurements, vertebral fracture risk assessment and blood and urine sampling for biochemical and immunological markers. T-cell and monocyte subsets were assessed using flow cytometry. Results: Phase One (n = 625): GFRFs were prevalent, but FRAX® scores and fragility fracture prevalence were low. Phase Two (n = 114): FRAX®-incorporated GFRFs and increased cumulative protease inhibitor exposure (but no other HIV disease-specific factor) were significant independent determinants of reduced BMD. Non-classical monocytes were also associated with reduced BMD. There was a significant negative correlation between FRAX® scores and BMD in black patients (p=0.003 for lumbar spine and total hip) and between FRAX® hip scores and total hip BMD in white patients (p=0.030). Total hip BMD differed significantly between patients with low FRAX® hip scores (0.999 ± .113 g cm-2) and high FRAX® hip scores (0.882 ± .136g cm-2) (p < 0.001). Conclusions: FRAX®-incorporated GFRFs were the predominant determinants of reduced BMD. FRAX® correlated well with BMD and may be of value for fracture risk assessment in specific HIV-positive patient subgroups.