During the course of 1 year, Clostridium difficile was isolated from 145 (4 per cent) of 3591 faecal specimens submitted to a hospital routine bacteriology laboratory. These C. difficile positive specimens came from 95 patients or 10.99 per cent of 864 patients investigated for suspected gastroenteritis. In a control population of 318 persons C. difficile was isolated from only 10 of them or 3.1 per cent; this difference was statistically significant. We conclude that as this increase in the isolation of C. difficile from patients with diarrhoea is significant and is about as frequent as that of Campylobacter spp., faecal specimens submitted for routine bacteriology should also be cultured for C. difficile . As with diarrhoea caused by other enteric pathogens, most C. difficile -associated episodes of diarrhoea are mild and spontaneously reversible but approximately 20 per cent of patients have prolonged diarrhoea and may benefit from treatment.