Over a 3.5 year period from August 1982 to December 1985, 200 Hickman catheters (Evermed, Medina, WA) were inserted into patients at the King Faisal Specialist Hospital and Research Centre (KFSHRC), Riyadh, Saudi Arabia. One hundred sixty catheters were placed in patients with malignant disease, many of whom were immunosuppressed at the time of catheter insertion. Seventy of 160 (44%) patients received prophylactic antibiotics and 90 (56%) did not. The mean age of each group was 23 years (range, 2 to 70 years), and the patients in each group were statistically similar in sex, underlying disease, and routine preoperative hematologic and biochemical evaluation. Exit-site wound infections occurred in 50 of 90 (55.5%) patients who did not receive prophylaxis and in 12 of 70 (17%) patients who received prophylaxis (P less than 0.0001). There was no statistically significant difference in the incidence of tunnel and incision site infections. The mean duration of antibiotic prophylaxis was 2.9 days (SD, 1.2 days). Organisms cultured from catheter associated infections, included Staphylococcus epidermidis 36, S. aureus 30, Klebsiella pneumoniae 1, Pseudomonas aeruginosa 3, Escherichia coli 1, and diphtheroids non-CDC-JK 3. Vancomycin was used as antibiotic prophylaxis in 64 patients, Kefzol (Eli Lilly, Indianapolis, IN) in one, oxacillin in three, nafcillin in one, and Septra (Burroughs Wellcome, Research Triangle Park, NC) in one. The data indicate that the use of intravenous antibiotic prophylaxis significantly reduces exit site infection and may reduce both tunnel and incision site infection. Prophylactic antibiotic coverage should be provided to patients during insertion of long-term indwelling right atrial catheters.