Rapid detection of MRSA may be important for the control of MRSA spread in hospitals. The aim of this investigation was to compare the use of a rapid polymerase chain reaction (PCR) screening method with standard culture for the detection of meticillin-resistant Staphylococcus aureus (MRSA) colonisation and to determine its impact on the incidence of MRSA in two hospital wards. During the first phase of the investigation (four months), patients in a surgical ward were screened using the rapid PCR technique and patients in a medical/cardiology ward were screened with standard culture methods. During the second phase of the investigation (four months), MRSA screening methods were switched between the two wards. An audit of infection control practices on each ward was made at the end of each phase in order to check whether any changes had occurred that might influence the risks of MRSA transmission. Use of the rapid PCR method significantly reduced the median time between swabs being taken, to the results being telephoned to the wards (excluding weekends), from 47 to 21 h (P<0.001). However, comparison of MRSA incidence during use of PCR (20/1000 bed-days) and culture methods (22.1/1000 bed-days) revealed no significant difference in incidence on the surgical ward (P=0.69). Regarding the medical/cardiology ward, analysis of data was complicated by an increase in the detection of MRSA during the PCR phase (P<0.05). The study demonstrated that rapid PCR can significantly reduce the turnaround times but reducing the time between swabs being taken to results being telephoned to the ward is still not sufficient to limit the transmission of MRSA.