Summary In 1957, a careful prospective study of the incidence of all surgical infections occurring on a surgical service at the Vancouver General Hospital was compiled, indicating at that time a high incidence of staphylococcal sepsis. Control of the epidemic was obtained by measures designed to eliminate staphylococcus from the ward and operating room and by reduction in the use of prophylactic antibiotics. A similar study in 1967 has confirmed the reduction in staphylococcal sepsis, and has shown an increase in coliform infection. The rate of clinical infection is nearly 20 per cent at present, as opposed to 30.1 per cent in 1957; however, only 18.1 per cent of the infections are due to Staph aureus as opposed to 44.4 per cent in 1957. The characteristic staphylococcus organism has changed, the multiple-resistant organism being uncommon (14.6 per cent in 1967 versus 55.5 per cent in 1957). Antibiotic usage has changed and appears to parallel changing sensitivity. Abandonment of prophylactic antibiotics has not resulted in preventable infection. Clean wound infection rate fell from 10.2 per cent in 1957 to 3.4 per cent in 1967, but roughly half of all clean wound infections continue to be due to Staph aureus. All major staphylococcal infection has been reduced from 9.9 to 2.0 per cent. Forty per cent of all such infections in 1967 occurred in the 7.9 per cent of patients still receiving prophylactic antibiotics. The increase in coliform infection has paralleled the increase in the use of broad spectrum antibiotics, currently ampicillin, to which many of the organisms responsible are resistant. It is possible that such an increase is nosocomial in origin, and its investigation requires that all infections be considered.