Clostridium difficile is a frequent and clinically important cause of diarrhoea that has been strongly but not exclusively associated with the hospital setting. The vast majority of cases of C. difficile diarrhoea are associated with antecedent treatment with antibiotics, of which cephalosporins and clindamycin appear to pose the highest risk. Within hospitals and chronic-care facilities, cross-infection of C. difficile has been related to transient carriage on hands of healthcare workers and contamination of diverse environmental surfaces, including electronic rectal thermometers. Prospective epidemiologic studies have demonstrated that acquisition of C. difficile is common in hospitalized patients. Although colonized patients contribute to nosocomial transmission of C. difficile, symptom-free carriage of C. difficile appears to reduce risk of subsequent development of C. difficile diarrhoea. Antimicrobial treatment with oral metronidazole or vancomycin to attempt to eradicate symptomless carriage is not recommended. Measures to control nosocomial C. difficile diarrhoea have focused on improved handwashing, use of barrier precautions with single rooms for symptomatic patients, reduction of environmental contamination, and antibiotic restriction. Restricting clindamycin has been particularly successful in terminating outbreaks of C. difficile diarrhoea associated with its use. The epidemiologic features of C. difficile and strategies for control are similar to those for micro-organisms that have acquired antimicrobial resistance. C. difficile may be indirectly or directly contributing to spread of resistant organisms, for instance, by causing diarrhoea and thereby enhancing environmental contamination with other gastrointestinal flora such as vancomycin-resistant enterococci. Thus, a consideration of C. difficile in the larger context of the world-wide spread of antibiotic resistance offers useful insights that may help form the basis for the development of more effective control measures.