We prospectively studied the difference in detection rates of multi-resistant Gram-positive and multi-resistant Gram-negative bacteria in the inanimate environment of patients harbouring these organisms. Up to 20 different locations around 190 patients were surveyed. Fifty-four patients were infected or colonized with methicillin-resistant Staphylococcus aureus (MRSA) or vancomycin-resistant enterococci (VRE) and 136 with multi-resistant Gram-negative bacteria. The environmental detection rate for MRSA or VRE was 24.7% (174/705 samples) compared with 4.9% (89/1827 samples) for multi-resistant Gram-negative bacteria (P<0.0001). Gram-positive bacteria were isolated more frequently than Gram-negatives from the hands of patients (P<0.0001) and hospital personnel (P=0.1145). Environmental contamination did not differ between the intensive care units (ICUs) and the general wards (GWs), which is noteworthy because our ICUs are routinely disinfected twice a day, whereas GWs are cleaned just once a day with detergent. Current guidelines for the prevention of spread of multi-resistant bacteria in the hospital setting do not distinguish between Gram-positive and Gram-negative isolates. Our results suggest that the inanimate environment serves as a secondary source for MRSA and VRE, but less so for Gram-negative bacteria. Thus, strict contact isolation in a single room with complete barrier precautions is recommended for MRSA or VRE; however, for multi-resistant Gram-negative bacteria, contact isolation with barrier precautions for close contact but without a single room seems sufficient. This benefits not only the patients, but also the hospital by removing some of the strain placed on already over-stretched resources.