Meticillin-resistant Staphylococcus aureus (MRSA) persists in the hospital environment and conventional cleaning procedures do not necessarily eliminate contamination. A prospective study was conducted on an intensive care unit to establish the level of environmental contamination with MRSA, assess the effectiveness of hydrogen peroxide vapour (HPV) decontamination and determine the rate of environmental recontamination. MRSA was isolated from 11.2% of environmental sites in the three months preceding the use of HPV and epidemiological typing revealed that the types circulating within the environment were similar to those colonising patients. After patient discharge and terminal cleaning using conventional methods, MRSA was isolated from five sites (17.2%). After HPV decontamination but before the readmission of patients, MRSA was not isolated from the environment. Twenty-four hours after readmitting patients, including two colonized with MRSA, the organism was isolated from five sites. The strains were indistinguishable from a strain with which a patient was colonized but were not all confined to the immediate vicinity of the colonized patient. In the eight weeks after the use of HPV, the environment was sampled on a weekly basis and MRSA was isolated from 16.3% sites. Hydrogen peroxide vapour is effective in eliminating bacteria from the environment but the rapid rate of recontamination suggests that it is not an effective means of maintaining low levels of environmental contamination in an open-plan intensive care unit.