Background And Objectives: Long-term central venous access is becoming an increasingly important component of health care today. Long-term central venous access is important therapeutically for a multitude of reasons, including the administration of chemotherapy, antibiotics, and total parenteral nutrition. Central venous access can be established in a variety of ways varying from catheters inserted at the bedside to surgically placed ports. Furthermore, in an effort to control costs, many traditionally inpatient therapies have moved to an outpatient setting. This raises many questions regarding catheter selection. Which catheter will result in the best outcome at the least cost? It has become apparent in our hospital that traditionally placed surgical catheters (ie, Hickmans and central venous ports) may no longer be the only options. The objective of this study was to explore the various modalities for establishing central venous access comparing indications, costs, and complications to guide the clinician in choosing the appropriate catheter with the best outcome at the least cost. Methods: We evaluated our institution’s central venous catheter use during a 3-year period from 1995 through 1997. Data was obtained retrospectively through chart review. In addition to demographic data, specific information regarding catheter type, placement technique, indications, complications, and catheter history were recorded. Cost data were obtained from several departments including surgery, radiology, nursing, anesthesia, pharmacy, and the hospital purchasing department. Results: During a 30-month period, 684 attempted central venous catheter insertions were identified, including 126 surgically placed central venous catheters, 264 peripherally inserted central catheters by the nursing service, and 294 radiologically inserted peripheral ports. Overall complications were rare but tended to be more severe in the surgical group. Relative cost differences between the groups were significant. Charges for peripherally inserted central catheters were $401 per procedure, compared with $3870 for radiologically placed peripheral ports and $3532 to $4296 for surgically placed catheters. Conclusions: Traditional surgically placed central catheters are increasingly being replaced by peripherally inserted central venous access devices. Significant cost savings and fewer severe complications can be realized by preferential use of peripherally inserted central catheters when clinically indicated. Cost savings may not be as significant when comparing radiologically placed versus surgically placed catheters. However, significant cost savings and fewer severe complications are associated with peripheral central venous access versus the surgical or radiologic approach. (AJIC Am J Infect Control 2001;29:32-40)