OBJECTIVE: To analyze a method that identifies potentially preventable nosocomial infections, as a tool to evaluate the performance of infection control programs through quantification of their potential for reducing nosocomial infections. METHODS: The database of the Study of the Prevalence of Nosocomial Infections in Spain (EPINE) was reanalyzed. The method was based on the use of false negatives of the classification table obtained from application of a fixed multiple logistic regression model, as an estimator of the number of potentially preventable nosocomial infections. RESULTS: The calculated number of patients with preventable infections was 7,493, which constituted 21.6% of the infected patients. Among hospital areas, intensive care had the lowest preventability rate (4.6%), whereas gynecology and obstetrics had the highest (40.6%). There was a significant inverse exposure-effect relationship between the proportion of preventable infections and the National Nosocomial Infections Surveillance (NNIS) System risk index. No correlation was observed between the prevalence of patients with nosocomial infection and the percentage of preventable infections. CONCLUSION: This analysis suggests that fewer nosocomial infections may be preventable in Spanish hospitals than previously assumed.