The efficacy of an educational intervention to prevent blood culture contamination (BCC) in internal medicine was studied in two medical wards in a busy tertiary-care hospital in which blood cultures were obtained by physicians rather than dedicated phlebotomists. Baseline BCC rates were 5.7% and 7.1% in intervention and control wards, respectively (p 0.6), compared with 1.95% and 6.7%, respectively, post-intervention (p < 0.001). Following multivariate analysis, only an absence of intervention was an independent variable associated with BCC. Thus simple educational intervention reduced BCC in internal medicine and was considered to be cost-effective.