At a regional hospital serving 150,000 inhabitants, the authors were tracking the occurrence of antibiotic-associated diarrhea caused by a nosocomial infection of the Clostridium difficile, especially its most serious form--pseudomembranous colitis. Six cases of the disease were found in a retrospective study in 2001, another 20 cases were diagnosed while actively searching for the disease from January till September of 2002. The patients suffered from many complications, were of an average age of 69.24 years, their hospitalization averaged 34 days and 15 (52%) of them underwent surgery during their hospitalization. It could be demonstrably proved that 25 cases of nosocomial infection occurred during hospitalization and 3 patients were admitted to hospital with the disease. However, 2 of them had been released from hospital less than 20 days before being admitted again. Twenty-five patients (96%) had been treated with antibiotics, often in combined therapy. Most often this involved penicillin with betalactamas activity (50%), clindamycin (42%) and cefalosporins (42%). We used methods for detecting enterotoxin A in the stool, rectoscopy and anaerobic cultivation of the stool in the diagnostic process. Rectoscopy discovered pseudomembranous colitis in 14 of 17 patients examined this way. The sensitivity for proving enterotoxin A in the stool using EIA, for patients with proven pseudomembranous colitis via rectoscopy, was 75%. Anaerobic cultivation of the stool was done in 12 patients and all the results were negative. The mortality rate of 38% for our group of patients testifies to the seriousness of this disease, which we consider to be the results of antibiotic therapy. In conclusion, nosocomial infection caused by Clostridium difficile is quite often a nosocomial disease, a prognosis that especially worsens for seriously ill patients.