This paper is a critical review of studies of educational programmes designed to improve prescription practices in ambulatory care. Scientific articles were selected from the following bibliographical indices: MEDLINE, IME, ICYT and ERIC. The searches covered the time period between 1988 and 1997. The search criteria included: primary-care, educat*, prescription* and other related keywords. The inclusion criteria were studies describing educational strategies aimed at general practitioners working in ambulatory settings. The study outcome was change in prescribing behaviour of physicians through prescribing indicators. The following data were extracted: study design, target drugs, type of intervention, follow-up period of the prescription trends, type of data analysis, type of statistical analysis and reported results. We found 3233 articles that met the search criteria. Of these, 51 met the inclusion criteria and 43 studied the efficacy/effectiveness of one or various interventions as compared to no intervention. Among seven studies evaluating active strategies, four reported positive results (57%), as opposed to three of the eight studies assessing passive strategies (38%). Among the 28 studies that tested reinforced active strategies, 16 reported positive results for all variables (57%). Eight studies were classified as a high degree of evidence (16%). We concluded that the results of our review suggest that the more personalized, the more effective the strategies are. We observe that combining active and passive strategies results in a decrease of the failure rate. Finally, better studies are still needed to enhance the efficacy and efficiency of prescribing practices.