OBJECTIVES:Cases of hypophosphataemia (often coincident with renal dysfunction) have been reported in HIV-infected patients taking tenofovir disoproxil fumarate (TDF), but randomized placebo-controlled trials of HIV-infected persons with normal baseline renal function have found a comparable incidence of hypophosphataemia in the TDF and placebo groups. We assessed the incidence of grade 2 and higher hypophosphataemia in the HIV Outpatient Study (HOPS). METHODS:We analysed a prospective cohort of patients who initiated either a TDF-containing highly active antiretroviral therapy (HAART) regimen [TDF-exposed (TDF+) group; n = 165] or a TDF-sparing HAART regimen [TDF-unexposed (TDF-) group; n = 90], and who had normal baseline phosphate and creatinine values. RESULTS:The TDF+ and TDF- groups had comparable median follow-up times (10.9 vs 8.8 months, respectively; P = 0.18) and number of phosphate measurements (median = 3 for both) and were similar on most clinical and demographic factors. During follow up, 12.7% of TDF+vs 6.7% of TDF-patients developed grade 2 hypophosphataemia (2.0-2.4 mg/dL), and 2.4% of TDF+ patients vs 0% of TDF-patients developed grade 3 hypophosphataemia (1.0-1.9 mg/dL); none developed grade 4 hypophosphataemia (<1.0 mg/dL). The incidence of grade 2 or higher hypophosphataemia was 16.7 per 100 person-years among TDF+ patients vs 8.0 per 100 person-years among TDF-patients (P = 0.11). CONCLUSIONS:The incidence of hypophosphataemia was somewhat elevated in HOPS patients who took TDF-containing HAART compared with those who took TDF-sparing HAART during the first 1 to 2 years of observation, but the difference was not statistically significant. Longer follow-up of a larger population is needed to determine if this trend towards an association achieves statistical significance and to evaluate the clinical consequences of hypophosphataemia.