Continuous ambulatory peritoneal dialysis (CAPD) has recently been introduced as an alternative to haemodialysis or peritioneal dialysis for patients suffering from chronic renal failure. The advantages of this procedure include the lack of need for special equipment in the home, the fact that bag changes can be carried out by the patient without assistance, the lack of need for vascular access, the increased mobility resulting from independence of a dialysis centre and an increased feeling of wellbeing.' However it has been found that a significant proportion of patients receiving this form of dialysis develop symptoms and signs suggestive of bacterial peritonitis.2 Conventional methods of culture of peritoneal dialysate from these patients have often proved negative despite the presence of pus cells in the peritoneal fluid and clinical evidence of infection.3 In theory there are three possible explanations for negative culture results. Firstly the condition may be non-bacterial in origin. Alternatively even if the infection is bacterial the causative agents may not be detected by the methods used because their numbers are small or because their cultural requirements are exacting. The third possibility is that bacteria may be prevented from growing in vitro by substances present in the dialysate. Various workers have attempted to improve the rate of recovery of bacteria.24 In one method 100 ml of dialysate is passed through a 0 45 ,um membrane filter-the filter is divided and half cultured aerobically, the other half anaerobically on enriched solid media. The disadvantages of methods involving filtration of large volumes of fluid include blockage of the filter with cells and contamination of filters during division and transfer to solid media. Another method involves inoculation of 2 ml of dialysate into nutrient broth and. chopped meat carbohydrate broth.