A total of 1144 patients receiving continuous ambulatory peritoneal dialysis in nine European centers was screened for nasal carriage of Staphylococcus aureus. Two hundred sixty-seven subjects were defined as carriers of S. aureus by having had at least two positive swab results from samples taken on separate occasions, and were randomly allocated to treatment or control groups. Members of each group used a nasal ointment twice daily for 5 consecutive days every 4 wk. The treatment group used calcium mupirocin 2% (Bactroban nasal; SmithKline Beecham, Welwyn Garden City, United Kingdom) and the control group used placebo ointment. Patients were followed-up for a maximum period of 18 months. There were 134 individuals in the mupirocin group, and 133 individuals acted as control subjects. There were no differences in demographic data, cause of renal failure, type of catheter, system used, or method of exit-site care between the groups. Similarly, there were no differences in patient outcome or incidence of adverse events between both groups. Nasal carriage fell to 10% in those subjects who received active treatment and 48% in those who used the placebo ointment. There were 55 exit-site infections in 1236 patient-months in the control group and 33 in 1390 patient-months in the treatment group (not significant). S. aureus caused 14 episodes of exit-site infection in the mupirocin group and 44 in the control group (P = 0.006, mixed effects Poisson regression model). There were no differences in the rate of tunnel infection or peritonitis. There was no evidence of a progressive increase in resistance to mupirocin with time. Regular use of nasal mupirocin in continuous ambulatory peritoneal dialysis patients who are nasal carriers of S. aureus significantly reduces the rate of exit-site infections that occurs because of this organism.