Twenty-five to 50 percent of all antibiotics prescribed for hospitalized patients are for prevention, not treatment, of infection. Procedures to institute rational, cost-effective utilization of these agents should have a significant impact on drug cost and pharmacy inventory. Several authors have described antibiotic cost reduction programs using pharmacy intervention. Unfortunately, measures that are successful in one institution may not be effective or appropriate in another. A three-year study was undertaken to examine the impact on physician prescribing of an infection control bulletin and formalized recommendation for antibiotic utilization. Patient records also were examined to determine if any change in antibiotic utilization would influence patient morbidity. Over the three examination periods there was a significant reduction in cost of prophylaxis in 7 surgical groups, and a trend toward cost reduction in 21 additional groups. Cost of prophylaxis also increased in other surgical groups. Cost reduction was associated with limited duration of prophylaxis and a shift toward use of first generation cephalosporin products. Patient morbidity did not differ significantly.