Mycobacterium fortuitum has been isolated from skin and soft tissue lesions with increasing frequency. Rarely, however, has it been a documented cause of peritonitis in patients receiving continuous ambulatory peritoneal dialysis. We report here the second such case and discuss both the possibility of M. fortuitum or similar organisms as one cause of "sterile" peritonitis in this patient population and the in vitro antimicrobial susceptibility testing of such isolates. Mycobacterium fortuitum has been isolated as a pathogen from several extrapulmonary sites. There has been, however, only one well-documented case of M. fortuitum causing peritonitis in a patient receiving continuous ambulatory peritoneal dialysis (CAPD) (13). We present here the second such case. The possible role of the organism in culturenegative peritonitis and methods of determining the in vitro antimicrobial susceptibilities of such isolates are discussed. Case report. A 32-year-old male being treated with CAPD for end-stage renal disease and with several medications for various psychiatric problems was admitted on 18 February 1985 with complaints of confusion and visual hallucinations. CAPD had been initiated in July 1984, and since that time the patient had experienced at least four episodes of peritonitis. On two occasions no organisms were cultured; Staphylococcus epidermidis and S. aureus were each cultured during two separate episodes. A physical examination at admission revealed a disoriented, thin male. His blood pressure was 134/90, his pulse was 88 beats per min and regular, and his temperature was 37.5°C. His abdomen was soft, and the catheter site was clean. The peripheral leukocyte count was 5,000/mm3, with a normal differential. The patient improved after psychotropic drugs were stopped, but he remained hospitalized to receive intensive psychotherapy. CAPD was continued. On 28 February 1985, the patient's temperature rose to 38.3°C, and he complained of abdominal pain. The peritoneal fluid was cloudy and contained 3,845 leukocytes per mm3, with 83% polymorphonuclear leukocytes. No organisms were revealed by Gram staining. The peripheral leukocyte count was 8,400 /mm3, with a slight left shift. Vancomycin and rifampin were administered after routine culturing of the peritoneal fluid was done. These cultures yielded no organisms. Fever and abdominal pain persisted, and the peritoneal fluid remained cloudy. The fluid was submitted for routine culturing as well as for culturing for fungi and mycobacteria on 3 March and 5 March. Routine cultures again showed no growth, and Gram staining revealed no organisms. Rifampin was discontinued, and gentamicin was begun. On 8 March, the peritoneal catheter was surgically removed, samples for culturing were taken of peritoneal fluid, peritoneal tissue, and the exit site, and a peritoneal biopsy was obtained. Histologic examination of the latter showed nonspecific chronic inflammation. On 9 March, the fluid submitted 6 days before was reported to be "positive for acid-fast bacilli,