OBJECTIVE: To evaluate the performance of the Rapid ID 32 Strep system in the hands of clinical microbiologists without expert knowledge of streptococci or enterococci. METHODS: One hundred and twenty-two strains of streptococci and enterococci conventionally identified in a reference laboratory were sent under code numbers to a clinical microbiology laboratory and identified with the Rapid ID 32 Strep system. RESULTS: Regardless of whether automatic reading and identification or visual reading with identification using tables were done, 75-77% of the 122 examined strains were correctly identified, 7% were misidentified and 16-18% could not be identified with certainty to the species level. The system correctly identified the majority of the examined pyogenic streptococci and enterococci, but only two-thirds of the viridans streptococcal strains. CONCLUSIONS: In a routine laboratory, the Rapid ID 32 Strep system can be used to give a rapid preliminary identification of streptococci and enterococci, but with viridans streptococci one would have to accept a certain risk of mis-identification. The assay can, however, be used to biotype viridans streptococci in order to attempt to establish identity between separate isolates, e.g. from blood in patients suspected of having endocarditis.