To differentiate imported and acquired strains of methicillin-resistant Staphylococcus aureus (MRSA), a 48-hour delay from hospital admission to the first MRSA-positive culture is usually considered. To assess if taking into account this delay without any other consideration is an accurate method, we defined 3 situations for whom we considered the MRSA acquisition status as questionable. The other situations were defined as either acquired MRSA or imported MRSA. We determined the acquisition status of MRSA (acquired, imported, or questionable) isolated during a 20-month period by considering or not considering screening samples performed on admission. The ratio "imported MRSA/acquired MRSA" (I/A) was calculated according to (1) the consideration of MRSA with questionable status as imported or acquired, and (2) the consideration of screening samples or not in the calculation of the ratio. The acquisition status in our hospital was questionable in 3.6% of patients when all samples were considered and in 12,0% when only clinical samples were taken into account (p = 0,01). The ratio I/A was 4-fold higher by considering both clinical and screening cultures and questionable status as imported than by considering only clinical samples and questionable status as acquired. Using a 48-hour delay without any other consideration is probably an accurate method to differentiate acquired and imported MRSA when a selective screening programme at admission in operational. Conversely, this definition seems to be more hazardous in the absence of screening.