International guidelines suggest that a high prevalence of meticillin-resistant Staphylococcus aureus (MRSA) infections should influence the use of vancomycin for surgical prophylaxis. In order to compare the efficacy and adverse effects of vancomycin versus cefazolin as antimicrobial prophylaxis for insertion of cerebrospinal fluid (CSF) shunts, a randomised prospective clinical trial was performed. Over a 16-month period, all consecutive adult patients who underwent CSF shunt insertion at a university hospital with a high prevalence of MRSA infections were included. Patients were randomly allocated to receive either vancomycin or cefazolin before surgery and followed-up for four weeks for the development of infections. Of the 176 patients included in the study, 88 received vancomycin and 88 cefazolin. Shunt infections were significantly less likely to be observed in patients who were on vancomycin prophylaxis (4% vs 14%; P=0.03). All isolated staphylococci were resistant to meticillin. Mortality of patients with post-surgical infections was higher in the cefazolin group (P=0.02). Our data suggest that use of vancomycin as prophylactic agent for cerebrospinal shunt placement reduces the rate of shunt infections in the context of high prevalence of MRSA.