This prospective, randomized study evaluated the risks associated with the Hickman catheter (HC) by a comparison of rates of placement and use complication, infection, treatment failure, and thrombogenicity with a group of patients with polyvinylchloride central venous catheters (CVC). Complications were studied prospectively. The patient populations of the two groups were similar and most patients were entered in the study because of prior venoablation. No significant difference in complication rates was noted for catheter placement complications, infection, or treatment failures. No catheter-related deaths occurred. Seven CVCs were exchanged for malfunction and one HC fractured and required repair. Venography was performed in seven patients in each group when the catheter was removed. The incidence of fibrin sheath formation was significantly increased in the HC group (P less than 0.05). One venous thrombosis was identified in a patient with a CVC. A separate group of patients with HCs who were not placed in the randomized study was followed. This group had a significantly higher incidence of septic complications than did the study group (P less than 0.005). We conclude that the HC is as safe as is the CVC when the conditions of placement and maintenance are rigorously controlled.