Experience of the diagnosis and epidemiology of Clostridium difficile in Sweden is reviewed. Samples from 5885 patients have been investigated at the National Bacteriological Laboratory in Stockholm from 1978–1983. Patients originate from all parts of the country and their number continues to increase. Cl. difficile seem to be of growing importance, especially in nosocomial infections. Most patients with antibiotic-associated diarrhoea and colitis (AAD/AAC) and Cl. difficile in their stools were above 60 years of age (63%) and there was a significant preponderance of females over males in the age groups 21–50 and above 60 years of age. The antibiotics most commonly associated with Cl. difficile enterocolitis (CDE) were penicillins, cephalosporins and clindamycin/lincomycin. On the basis of consumption clindamycin/lincomycin and cephalosporins are associated 70 and 40 times, respectively, more often than penicillins in CDE. Diagnosis of CDE relies mainly on detection of the cytotoxin (toxin B) in stool specimens. It was present in 873/4793 (18.2%) patients whereas the bacterium was found in only 12%. An immunoassay for detection of the enterotoxin (toxin A) of Cl. difficile seems to be a useful alternative to the cytotoxin assay, but some stool specimens with a low toxin B titre were negative. Five specimens negative for toxin B were positive for toxin A and came from patients where additional information suggested CDE. A serological assay for demonstration of circulating antibodies to Cl. difficile toxins has also been evaluated and is positive in about half of the cases of CDE. Antibody response seems to be absent or delayed in patients with relapse of colitis after antibiotic treatment.