Between October and December 2000, a region-wide prevalence study of hospital-acquired infections (HAI) was conducted in all public hospitals (59 facilities with ca. 16000 beds; 560000 admission yearly) in Piemonte Region, Italy, and in the one hospital of the neighbouring autonomous region of Valle d'Aosta. The study population comprised a total of 9467 patients hospitalized for at least 24 h. The prevalence of HAI was 7.84%, with marked differences in prevalence among the participating hospitals (range: 0-47.8%). The higher relative frequency of urinary tract infections (UTI; 52.7%) was due to the inclusion of urine cultures obtained on the day of the study from asymptomatic UTI in catheterized patients. A significant correlation was found with major risk factors related to medical procedures (urinary catheter, mechanical ventilation, surgical drainage, intravascular catheters). Patients with HAI were found to be older and to have a greater mean length of stay in hospital. Multiple logistic regression analyses showed that lack of independence, indwelling urinary catheter and mechanical ventilation were the risk factors more significantly associated with HAI. The use of antibiotics, in particular prophylactic agents used in surgery (cephalosporins, glycopeptides), provided an incentive for corrective intervention in antibiotic administration and in training of healthcare workers.