Between April 1994 and May 1995 4000 adult patients admitted to selected specialties of a district general hospital were recruited to this study. Hospital-acquired infections presenting during the in-patient stay were identified using previously validated methods of surveillance, and information on daily resource use by both infected and uninfected patients was recorded and estimates of their cost derived. Linear regression modelling techniques were used to estimate how much of the observed variation in resource use and costs could be explained by the presence of an infection. Complete in-patient data sets were obtained for 3980 patients. Of these, 309 patients (7.8%; 95% CI; 7.0, 8.6) presented with one or more hospital-acquired infections during the in-patient period. Infected patients, on average, incurred hospital costs 2.9 (regression model estimate: 2.8; 95% CI; 2.6, 3.0) times higher than uninfected patients, equivalent to an additional pound3154 (regression model estimate pound2917). Both the incidence and the economic impact varied with site of infection and with admission specialty. Estimates of the burden of hospital-acquired infections occurring in adult patients admitted to similar specialties at NHS hospitals in England were derived from the results of this study. An estimated 320 994 (95% CI; 288 071, 353 916) patients per annum acquire one or more infections which present during the in-patient period, and these infections cost the hospital sector an estimated 930.62 million pounds (95% CI; 780.26 pounds; 1080.97 million pounds) per annum. The results presented represent the gross economic benefits that might accrue if these infections are prevented. Further research is required to establish the net benefits of prevention.