Subtyping isolates may be useful for epidemiological studies of methicillin-resistant-Staphylococcus aureus (MRSA) outbreaks. Among subtyping methods, DNA-based techniques have been applied very effectively for this purpose. An outbreak of MRSA infections took place in one hospital in Barcelona early during 1991. From the beginning of the outbreak to December 92, 70 MRSA isolates from different patients and sources were collected. All strains were evaluated by restriction endonuclease analysis of plasmid DNA (REAP) and macrorestriction endonuclease analysis of genomic DNA using Sma I and pulsed-field-gel-electrophoresis (PFGE). Plasmid screening and REAP using Hind III demonstrated two plasmid subtypes: subtype A showing a large plasmid, and subtype B showing the same large plasmid plus a smaller one. Subtypes A and B corresponded to the more recent and older isolates, respectively, suggesting the loss of the small plasmid during the epidemic. PFGE using Sma I displayed two closely related profiles (PFGE subtype A and A'; CS = 0.90). These subtypes were different from those subtypes exhibited from 4 methicillin-susceptible-Staphylococcus aureus (MSSA) isolates from the same hospital and from 2 epidemiologically unrelated MRSA isolates. Almost all isolates showing PFGE subtype A preceded those isolates showing PFGE subtype A'. This fact and the similarity between both subtypes suggested minor chromosomal DNA rearrangement during the outbreak from a unique strain. While PFGE using Sma I is a useful tool in evaluation of clonal dissemination, our data suggest epidemic or local outbreaks may need several methods to best delineate the source and spread of MRSA strains. The reproducibility and discriminatory power of REAP makes it a useful adjunct in this context.