We contrast the experiences, in our Health Authority in South-East London, with the particular epidemic methicillin-resistant Staphylococcus aureus ( the EMRSA) strain that has recently spread widely around London and South-East England, and with the other MRS A (OMRSA) strains encountered there. Our isolates of the EMRSA were identical by chromosomal restriction enzyme analysis, and the chromosomal and plasmid phenotypes were similar to those described in North London and Eastern Australia. Experimental phage-typing distinguished them from OMRSA encountered in 1984 to 1986. Between 1984 and 1985, the EMRSA caused increased infection and patient colonization compared to the years 1969 to 1983. A change in infection control procedures was usually required to control the EMRSA and in 1986 isolates had returned to their pre-1984 levels. Between 1984 and 1986 OMRSA were still encountered, but did not spread or require changes in infection control procedures. The distribution of other resistant isolates was examined; c 94% of neomycin-resistant isolates were in-patients or clinic patients. Forty-five different phage-type/antibiogram patterns were found in 88 isolates from 66 patients between 1982 and 1985, and patient clusters were uncommon. The ability of the EMRSA to spread is discussed and is probably not purely organism related. Our experience supports the contention that some MRSA are truly epidemic, whilst others do not behave in this manner.