The choice of initial antibiotic therapy for continuous ambulatory peritoneal dialysis (CAPD) peritonitis is complex. To test the hypothesis that the initial choice of intraperitoneal antibiotics does not influence the outcome of therapy for CAPD peritonitis, we studied 1930 patients on peritoneal dialysis as of 1 January 1991, and recorded all episodes of peritonitis for one year. There were 1168 episodes of peritonitis for which 64% of the patients received intraperitoneal antibiotics as the first course of treatment. We determined peritonitis resolution, technique failure, hospitalization, and catheter removal rates as measures of successful treatment for peritonitis. Cephalothin, cefazolin, vancomycin, and ceftazidime were used alone or in combination with the aminoglycosides, tobramycin and gentamicin. The mean resolution rate was 65% following the first course of antibiotics and 76% by the third course of antibiotics, if necessary. The antibiotic regimens were not different with respect to resolution rate except when aminoglycosides were used alone. Technique failure was also greater in patients treated with aminoglycosides without other antibiotics. Hospitalizations were increased in patients treated with single drug therapy. If initial intraperitoneal therapy includes treatment for both gram-positive and gram-negative bacteria, the initial choice of intraperitoneal antibiotics does not influence the outcome of therapy for CAPD peritonitis.