On 2 November 1999, one of the main hospital façades adjoining cardiovascular surgery collapsed in a 900-bed teaching hospital in Santander, Spain. The purpose of this study was to determine whether the accident affected the safety of patients by increasing the risk for nosocomial and surgical site infections (SSI). Measures for the prevention of nosocomial infections were immediately reinforced. A total of 217 consecutive patients were operated on before 2 November 1999, with another 296 after this date. Patients in both study periods showed similar severity of illness, complexity of surgical procedure and length of hospital stay. The overall rate of nosocomial infection before and after the accident was 28.1% and 24.7%, respectively (P=0.381). The rates of respiratory infection, urinary infection and bacteraemia were also similar. A statistically significant reduction in the SSI rate in the second period was observed (14.8% vs 4.4%, P=0.008). The collapse of the façade was not associated with any increase in nosocomial infection rates, but there was a significant reduction of SSI rates in relation to intensive infection control measures implemented after the collapse.