The study objective was to measure the benefits of elimination of nasal carriage of Staphylococcus aureus by calcium mupirocin ointment in patients undergoing continuous ambulatory peritoneal dialysis. The design was a prospective, placebo-controlled, randomized clinical trial. The subjects were 267 patients recruited from nine renal units in Belgium, France and the UK. The main outcome measures were the rate of catheter exit site infection (ESI), rates of other infections and healthcare costs from the perspective of a hospital budget-holder. The rate of ESI caused by S. aureus was significantly reduced from one in 28.1 patient months to one in 99.3 patient months (P = 0.006) and there were also non-significant trends towards lower rates of ESI caused by any organism and peritonitis caused by S. aureus. In comparison with the placebo group, patients in the mupirocin group with ESI had lower antibiotic (P = 0.02) and hospitalization costs (P = 0.065). However, overall costs of antibiotic treatment, for all infections combined, were not significantly different (P = 0.2) and total antibiotic costs (including mupirocin) were significantly higher in the mupirocin group (P = 0.001). Mupirocin prophylaxis would have been cost-neutral if the rate of ESI increased to >75% in the placebo group, or if all healthcare costs increased by 40%, or if the cost of screening was reduced from Pound Sterling 15 to Pound Sterling 3 per patient, or if the cost of mupirocin treatment was reduced from Pound Sterling 93 to Pound Sterling 40 per patient year. In conclusion, savings in healthcare costs are unlikely to be sufficiently great to offset the cost of mupirocin and screening for nasal carriage of S. aureus. The decision about whether or not to implement mupirocin should depend on a local analysis of the value of preventing ESIs caused by S. aureus.