Intraperitoneal antibiotics are used to treat Pseudomonas aeruginosa peritonitis, a serious complication of continuous ambulatory peritoneal dialysis. However, P. aeruginosa killing is often inefficient despite low MBCs. Broth dilution MIC/MBC and time kill curves of tobramycin, amikacin, netilmicin, azlocillin, piperacillin, ceftazidime, cefsulodin, and ciprofloxacin were determined in peritoneal dialysis fluid (PDF), buffered PDF, fluid recovered from patients on continuous ambulatory peritoneal dialysis (RPF), and cation-supplemented Mueller-Hinton broth. MBCs of all antibiotics were 8 to 16 times greater in PDF and RPF than in Mueller-Hinton broth or buffered PDF. Use of the time kill curve technique and Mueller-Hinton broth showed that aminoglycosides killed greater than or equal to 99.9% of P. aeruginosa at 1 h, ciprofloxacin killed greater than or equal to 99.9% at 2 h, and beta-lactams killed greater than or equal to 99.9% at 6 h. In contrast, killing was not demonstrated in PDF by any drug at 6 h and by aminoglycosides only at 24 h. Bactericidal activity was optimal in RPF for ciprofloxacin at 1 h and for aminoglycosides at 2 h; bactericidal activity was not demonstrated in RPF with any beta-lactam (no kill by penicillins; less than 99% kill by cephalosporins). Slow bacterial growth, increased protein binding, and glucose concentrations and other inhibitors may interfere with beta-lactam activity in RPF. These considerations and reported clinical failures and toxicity of aminoglycoside therapy warrant further study of quinolones and drug combinations in P. aeruginosa peritonitis.