Infection remains a major complication of central veinous catheters. A clinically or bacteriologically proven infection is identified in 5 to 10% of the cases although the actual risk varies from less than 1 to more than 30 episodes of infection per 1000 catheter-days. Systemic spread follows colonization of the catheter tip via the skin or the connectors, less frequently via hematogenous seeding from a distant focus of infection. Excepting emergency situations, search for proof of catheter-related infection or noninfection without removing the catheter now appears to be acceptable for selected intensive care patients if appropriate safety measures are taken; the reasonable choice is between maintaining the catheter while performing tests susceptible of determining the reality of catheter-related infection (in this case there remains a certain doubt concerning the respective performance of the different methods proposed for early diagnosis) and catheter exchange on a metal or plastic guide-wire. Rigorous application of the rules of hygiene, asepsis, and local care is crucial: systematic and repeated hand washing is by far the best antiinfectious prophylaxis. Three types of future developments susceptible of improving the current situation can be identified: better knowledge of host defense mechanisms against foreign bodies, reduction of hand-transmitted contamination by wider use of hydro-alcohol solutions for decontamination of the hands, and elaboration of new connectors, and most importantly, biomaterials capable of inhibiting colonization by the growing number of increasingly virulent strains. Discussions concerning the real clinical efficiency, ecological risk and the cost-effectiveness of catheters impregnated with antiseptics and/or antibiotics is a perfect illustration of the current debate.