Abstract: In recent years, various studies highlighted the key role of renal impairment (RI) as a cause of morbidity and mortality in HIV-1 infected patients likely favored by the increased patients’ longevity, concomitant comorbidities, ART toxicity, polypharmacy. So far, no study has focused yet on the possibility to early detect people living with HIV (PLWH) that will develop a renal impairment after antiretroviral therapy (ART) start by use of potent mathematical models based on machine learning systems. This possibility would be helpful in adopting strategies for prevention of RI. Materials and Methods We examined 1212 HIV-1 infected patients, followed at the Infectious Diseases Department of the San Raffaele Scientific Institute, who started ART between January 1998 and December 2014, with glomerular filtration rate (eGFR) determinations available both at the start of ART (baseline, BL) and ≥2 determinations over the 5 years after BL. eGFR was calculated using the CKD Epidemiology Collaboration Equation (CKD-EPI). A diagnosis of renal impairment (RI) was defined by the occurrence of two consecutive eGFR determinations