Background Peritoneal dialysis (PD)-related infections are the major cause of technique failure. Exit-site infections (ESI) can be prevented by local application of antibiotics. Mupirocin (M) is the most extensively studied drug for this application. Long-term use can result in the development of resistance. Gentamicin (G) is an attractive alternative, with both gram-positive and gram-negative activities. We studied the comparative efficacy of G cream versus M ointment in the prevention of PD-related infections in a Chinese cohort. Methods This was a prospective study of adult PD patients of the Princess Margaret Hospital, Hong Kong. Patients were excluded if they had active infection, recent ESI or peritonitis, history of allergy to either drug, or were unable to apply the drug or give consent. Patients were taught to apply the drug daily to the exit site after routine exit-site care. Records were tracked prospectively during hospital admissions and clinic follow-ups. Results 95 patients were recruited; 14 discontinued the study. The ESI rates were 0.38 and 0.20 episodes/patient-year for the G group and the M group respectively ( p = 0.36). Gram-positive ESI rates were 0.18 and 0 episodes/patient-year for the G group and the M group respectively. Gram-negative ESI rates were 0.20 episodes/patient-year for both groups ( p = 0.62). The overall peritonitis rates were similar in the two groups ( p = 0.91). Discussion In addition to good perioperative care and strict exit-site care, local antibiotic application can prevent ESI. Mupirocin has been extensively studied and shown to be effective. Similar if not superior effects of G cream have been demonstrated. In this study, neither antibiotic gave significantly better results in the prevention of either ESI or peritonitis. Conclusions Both gentamicin and mupirocin were effective as prophylaxis for ESI. Longer study is required to determine the long-term efficacy and the potential beneficial effect on the prevention of peritonitis.