Central venous catheters facilitate venous access, allowing the intravenous administration of complex drug treatments, blood products and nutritional support, without the trauma associated with repeated venepuncture. However, central venous catheters are associated with a risk of infection. Some studies have indicated that the type of dressing used for central venous catheters may affect the risk of infection. Gauze and tape or transparent polyurethane film dressings such as Tegaderm, Opsite or Opsite IV3000 are the most common types of dressing used to secure central venous catheters. Currently, it is not clear which type of dressing is the most appropriate. To compare gauze and tape and transparent polyurethane central venous catheter dressings in terms of catheter related infection, catheter security, tolerance to dressing material and dressing condition in hospitalised adults and children. The Cochrane Wounds Group Specialised Trials Register (October 2002), the Cochrane Controlled Trials Register (4th Quarter 2002) and the databases; MEDLINE (1966-December 2002, CINAHL (1982-October 2002) and EMBASE (1980-December 2002) were searched to identify any randomised controlled trials comparing the effects of gauze and tape and/or transparent polyurethane dressings for central venous catheter sites. Additional references were identified from bibliographies of published literature and were also sought from other sources. All randomised controlled trials evaluating the effects of dressing type (i.e. gauze and tape and/or transparent polyurethane dressings) on central venous catheter related infection, catheter security, tolerance to dressing material and dressing condition in hospitalised patients. Twenty-three studies were reviewed. Data was extracted from each paper by two members of the review team independently and results then compared. Differences were resolved either by consensus or by referral to a third member of the review team. Authors were contacted for missing information. Of the 23 studies reviewed, 14 were excluded. Nine studies were included. Data was only available for meta-analysis from six of the nine included studies. Of the six included studies with available data, two compared gauze and tape with Opsite IV3000, two compared Opsite with Opsite IV3000, one compared gauze and tape with Tegaderm, and one compared Tegaderm with Opsite. There was no evidence of any difference in the incidence of infectious complications between any of the dressing types compared in this review. Each of these comparisons was based on no more than two studies and all of these studies reported data from a small patient sample. Therefore it is probable that the finding of no difference between dressing types is due to the lack of adequate data. There is a high level of uncertainty regarding the risk of infection with the central venous catheter dressings identified in this review. Therefore, at this stage it appears that the choice of dressing for central venous catheters can be based on patient preference. To identify the most appropriate central venous catheter dressings, further research is necessary. It is paramount that any future studies investigating this issue must be rigorously performed randomised controlled trials.