The aim of this study was to assess to what extent patients with meticillin-resistant Staphylococcus aureus (MRSA) at respiratory sites shed viable MRSA into the air of hospital rooms. We also evaluated whether the distance from the patient could influence the level of contamination. Air sampling was performed directly onto MRSA-selective agar in 24 hospital rooms containing patients with MRSA colonization or infection of the respiratory tract. Samplings were performed in duplicate at 0.5, 1 and 2-3 m from the patients' heads. Clinical and environmental isolates were compared using antimicrobial resistance patterns and pulsed-field gel electrophoresis. MRSA strains were isolated from 21 out of 24 rooms, in quantities varying from between 1 and 78 cfu/m3. In each of the 21 rooms, at least one of the environmental isolates was identical to a clinical isolate from the patient in that room. There was no significant difference in MRSA counts between the distance from the patient's head and the sampler. This study demonstrates that most patients with MRSA infection or colonisation of the respiratory tract shed viable MRSA into the air of their room. The results emphasise the need to study MRSA in air in more detail in order to improve infection control recommendations.