Two microbiology laboratories, one serving an inner city hospital and one a rural public health laboratory, collected data on the outcome of examining faecal specimens in 1994. Overall, 6.7% of the investigations were positive, but the rates were lower for hospital inpatients, for recently described pathogens, and in the absence of relevant clinical details; rates were higher for patients with a history of foreign travel. No benefit was gained by examining more than two specimens from any patient. Clostridium difficile was the only investigation frequently positive among patients already in hospital, and virological tests were often positive in patients investigated by general practitioners.