INTRODUCTION: The mechanisms underlying the effect of tenofovir disoproxil fumarate (TDF) on the decline of bone mineral density (BMD) have not been established, especially the effect of renal tubular dysfunction. METHODS: Longitudinal study of 90 patients with two successive dual X-ray absorptiometry scans after evaluation of serum and urinary parameters (proteinuria, albuminuria, phosphaturia, uricosuria, glycosuria, ß-2-microglobulin, and retinol-binding protein). RESULTS: After a median of 38 months on TDF, osteopenia at spine and hip was observed in 49 and 48%, and osteoporosis in 9 and 2%, respectively. There was a lineal correlation between BMD at femoral neck and time on TDF (Spearman's rho = -0.27; P = 0.01). One or more tubular abnormalities were observed in 80% of cases (hyperphosphaturia, 50%). A lower BMD correlated with phosphaturia (r = -0.25; P = 0.03), even with phosphataemia within normal limits. In fact, patients with previous improvement in phosphaturia had better BMD at inclusion (Spearman's rho = -0.33; P < 0.01). A second dual X-ray absorptiometry, after a median of 40.8 months (33.8-45.1; 627.7 patients-year on TDF), showed additional BMD reduction at hip in 50% of cases (36% with bone loss >3%), a decline associated with phosphaturia (ß, -0.31; P = 0.01) or number of tubular abnormalities (ß, -0.41; P = 0.01), but also with use of boosted protease inhibitors (ß, -0.47; P = 0.03) and BMD at inclusion (ß, -0.33; P = 0.03). CONCLUSION: Chronic abnormal phosphaturia explains, at least in part, progressive bone loss during TDF therapy. These data suggest that tubular dysfunction leads to an altered equilibrium between phosphataemia, phosphaturia, and bone as mechanism of progressive BMD decline.