Background A clinical trial comparing the rate of discontinuation and tolerability of two post-exposure prophylaxis (PEP) regimens was performed. Methods A total of 255 individuals attending the emergency rooms of six hospitals for exposure to HIV and criteria to receive PEP were randomized to receive zidovudine/ lamivudine plus either lopinavir/ritonavir ( n=131) or atazanavir ( n=124; day 0). The primary end point was the rate of PEP discontinuation before day 28 of follow-up. Secondary end points were incidence of side effects, follow-up at days 90 and 180 and rate of seroconversions. Results A total of 55 patients (29 in lopinavir/ritonavir and 26 in atazanavir arms) did not attend the first scheduled appointment (day 1) and were excluded from the analysis. The rate of discontinuation before day 28 owing to any cause was similar between groups (37/102 [36%] in lopinavir/ritonavir and 35/98 [36%] in atazanavir arms, P=0.82). Adverse events were the reason for discontinuation or switching of PEP in 33 individuals (16/102 [16%] in the lopinavir/ritonavir arm and 17/98 [17%] in the atazanavir arm, P=0.84). Adverse events were reported in 92/200 (46%) of individuals on PEP who attend at least the day 1 appointment (50/102 [49%] in the lopinavir/ ritonavir arm and 42/98 [43%] in the atazanavir arm, P=0.38). There were no seroconversions. Conclusions The rate of discontinuation of PEP before day 28 was similar with zidovudine/lamivudine plus either lopinavir/ritonavir or atazanavir. The rate of discontinuation of PEP because of adverse events was low in both arms. Almost 50% of the patients of both arms suffered side effects. New strategies are needed to improve the tolerance.