Background To investigate change in renal function in African patients initiating second-line antiretroviral therapy (ART) including ritonavir-boosted protease inhibitor (PI/r) with or without tenofovir disoproxil fumarate (TDF). Methods HIV-1-positive adults, failing standard first-line ART were randomized to either TDF/emtricitabine (FTC)+LPV/r, abacavir + didanosine +LPV/r or TDF/FTC+ darunavir (DRV)/r and followed for 18 months. Patients with an estimated glomerular filtration rate (eGFR) ≥60 ml/min/1.73 m² at baseline were included in this analysis. Results Data from 438 out of 454 randomized patients were analysed. Median age was 38 years and 72% were women. Initiation of PI/r-based second-line regimen induced a marked eGFR decline of -10.5 ml/min/1.73 m² at week 4 in all treatment groups with a greater decrease in TDF/FTC+LPV/r arm (-15.1 ml/min/1.73 m²). At month 18, mean eGFR in the non-TDF containing regimen recovered its baseline level and was significantly greater than eGFR 18-month levels in the TDF-containing regimens that experienced only partial recovery (difference: -10.7; CI -16.8, -4.6; P=0.001 in TDF/FTC+LPV/r and -6.4; CI -12.5, -0.3; P=0.04 in TDF/FTC+DRV/r). At 18 months, prevalence of stage 3 chronic kidney disease was low (<3%) and not associated with treatment. One treatment discontinuation and five TDF dosage reductions for renal toxicities were reported in TDF-containing arms. Conclusions Overall, these results suggest a reasonable renal tolerance of a regimen associating TDF/FTC+PI/r in African patients with eGFR>60 ml/ml/1.73 m² at baseline. They also support the recommendation of reassessing renal function 1 month after initiation of treatment including ritonavir to account for the ritonavir-related artefactual decrease of eGFR and determine the new reference baseline value.