This paper aims to evaluate the effectiveness of a program designed to improve the quality of drug prescription-writing at a university hospital in France. Improvement actions included feed-back from yearly audits and the dissemination of recommendations on how best to write the prescriptions. A random sample of 30 stays was selected from among the hospitalizations for the year 1996. From each patient, medical records were searched for the first prescription order of the stay and its quality was assessed according to standards. A total of 872 records were relevant and included 3,289 medications. The results were compared to those obtained for the two previous years. Actions to sensitize prescribers resulted in an insufficient improvement of most indicators of prescription-writing quality with results remaining well below ideal standards. The hospital staff concerned had a positive opinion of the program which led to an awareness of prescription problems. This assessment showed that the program had a moderate impact on prescribers' practice and efforts must be continued.