Efficacy and tolerability of nevirapine-based HAART regimens were retrospectively evaluated. HIV-1-infected patients were included if they had been receiving a NVP-containing HAART regimen for at least 60 months, regardless of the reason for its initiation. A total of 82 patients were included in this study. The median follow-up time period was 96 months (range 64-120). At the time of starting NVP-based therapy, 31.7% (26 patients) were antiretroviral-naive, while 68.3% (56 patients) switched to an NVP-based strategy because of intolerance to the prior regimen, failure of a previous regimen or for simplification purposes. Coinfection with hepatitis B or C viruses was present in 26.8% of the patients (n 22). The most frequent nucleoside analogue backbone was zidovudine/lamivudine (52.4%) followed by stavudine/lamivudine (15.8%). A protease inhibitor was concomitantly administered in 2.4% of cases. To investigate the tolerability, we report the results separately on the basis of the presence/absence of HBV and/or HCV coinfection. Coinfected patients displayed significantly higher baseline liver enzymes than non-coinfected patients. Nevertheless, ALT levels showed a constant decrease over time in coinfected patients. Overall, median triglycerides, HDL and total cholesterol values tended to be lower in coinfected patients than in the remaining subjects. In particular, there was a significant decrease in triglyceride values in both groups and, concomitantly, a slight increase in total cholesterol. On the other hand, HDL cholesterol demonstrated a progressive increase. Finally, no change was found in glucose plasma levels, which always remained within normal ranges. In conclusion, no long-term toxicities associated with the use of nevirapine beyond five years were documented in our cohort of patients, even in coinfected patients. Long-term exposure to nevirapine was associated with optimal HIV suppression and favourable lipidic and glucidic profiles.