To compare the expenditure and usage of antibiotics at the San Martino Teaching Hospital, a 2500-bed hospital in Genoa, Italy, before and after the implementation of an antibiotic control program that streamlined the hospital formulary and the creation of a restricted group of antibiotics requiring approval before use. Usage and expenditure data for all antibiotics were collected from 1996 to 1998. Antibiotic usage was standardised by defined daily doses (DDDs) per 100/patient-days. Cost data were expressed in Euros. Changes in antibiotic usage was determined by comparing the mean usage during 1996 and 1997, the period before the implementation of the antibiotic control program, to 1998 when the streamlined formulary and restricted group of antibiotics, controlled by the Infectious Disease Team (IDT). were initiated. The Wilcoxon rank sign test was used to determine statistical significance of the changes in overall antibiotic use; a P value of less than 0.05 was considered significant. After the implementation of the antibiotic control program, overall antibiotic usage decreased by 8.5%, 28.00 DDD/100 patient-days during 1996-1997 to 25.62 DDD/100 patient-days during 1998. The control program resulted in overall savings of 342,927 Euros after the first year of implementation. The usage and expenditure in the restricted group of antibiotics decreased by 78.5% and 53.5%, respectively, (P=0.03). Restricting the use of ceftazidime and imipenem accounted for the majority of the decreased usage and savings. In the non-restricted group of antibiotics, usage increased only by 32.6% resulting in a net reduction of 46.3% in all antibiotic use. Although antibiotic control programs have been successful in other countries, this represents the first attempt at successful antibiotic control in a large Italian teaching hospital. Streamlining the formulary to control antibiotic choices and the creation of a restriction program using the expertise of infectious disease physicians resulted in significant reductions in the use of and expenditure for antibiotics.