Background: Colonization pressure has been confirmed as an important risk factor for methicillin-resistant Staphylococcus aureus (MRSA) acquisition among inpatients, but their precise relationship has not been well investigated. Because MRSA carriers can disperse MRSA into their immediate environment with different abilities, the relationship among colonization pressure, the degree of MRSA contamination in environment surrounding MRSA carriers, and MRSA transmission should be explored to facilitate efficient implementation of infection control measures. Methods: Active MRSA screening and environmental sampling were performed in a 23-bed emergency ward (EW) and a 7-bed respiratory intensive care unit (RICU) between March 2009 and February 2010. Weekly colonization pressure (WCP) was modified to WCPe (WCP adjusted by degree of environmental contamination). Receiver operating characteristic curve and correlation analyses were used to analyze the accuracy of WCPe in predicting MRSA acquisition and their correlation, respectively. Results: We found that 34.1% (858/2,520) of the immediate environmental sites of MRSA-positive patients were contaminated with MRSA. The areas under the receiver operating characteristic curve of WCPe reached as high as 0.784 (95% confidence interval, 0.659-0.909; P < .01) for the EW and 0.866 (95% confidence interval, 0.766-0.967; P < .01) for the RICU. Spearman's correlation coefficient (r) indicated a positive and significant correlation between WCPe and MRSA acquisition rate in the subsequent weeks for both the EW (r = 0.45; P = .001) and RICU (r = 0.51; P < .001). Conclusion: Active MRSA screening combined with targeted environmental contamination monitoring could be a more efficient measure for determining the magnitude of the MRSA reservoir in wards occupied by MRSA carriers. WCPe showed moderate prediction accuracy for both the EW and the RICU, and a threshold WCPe value may be used as a predictor to enhance infection control measures, especially for medical facilities without a sufficient number of single rooms.