In a prospective randomized clinical trial 84 peritonitis episodes were treated with cephradine, either orally or intraperitoneally. No difference in treatment outcome between both groups could be demonstrated. In episodes caused by susceptible micro-organisms a good response was seen in 82% in the oral and 82% in the intraperitoneal groups. These clinical findings were supported by the demonstration of adequate cephradine concentrations in serum and dialysate after oral as well as after intraperitoneal administration. Altogether cephradine was given orally or intraperitoneally in 88 episodes of peritonitis as drug of first choice. In 52 a complete cure was obtained, in 36 another antibiotic was subsequently needed as soon as bacterial susceptibility was known. No patient deteriorated appreciably during the delay between the start of cephradine and the switch to another antibiotic. Of the 36 episodes 14, caused by methicillin-resistant Staphylococcus epidermidis, responded well initially to cephradine but relapsed later. Change to another antibiotic effected a complete recovery in all 14 cases. Of the remaining 22 episodes, 14 were cured by the other antibiotic, in eight the catheter had to be removed. Aminoglycosides could be avoided except for ten of the episodes. During peritonitis CAPD was continued, in 71% of the cases on an outpatient basis. Mortality due to peritonitis was absent. We conclude that oral cephradine can be used as drug of first choice in the initial treatment of CAPD peritonitis, because a good initial response was obtained in 66 (52 + 14) i.e. 75% of 88 episodes. However, complete cure by cephradine alone was achieved in only 60%.(ABSTRACT TRUNCATED AT 250 WORDS)