BACKGROUND: Vibrio vulnificus sepsis requires a rapid and accurate bacteriological diagnosis for optimal management of the patient because of its high mortality. We evaluated two automated bacteriological identification systems, Microscan (WalkAway 96, Dade Behring, West Sacramento, CA, USA) and Vitek II (bioMerieux, Durham, NC, USA), for their ability to identify V. vulnificus strains isolated from clinical specimens. METHODS: A total of 60 V. vulnificus strains isolated from clinical specimens in Chonnam University Hospital during 1993-2003 were tested. For the identification of the isolates by the Microscan, Neg Combo type 32 was used and four different panel inoculation methods were evaluated for accuracy. Identification by the Vitek II system was carried out using Vitek ID-GNB cards in accordance with the manufacturers, instruction using 0.45% saline media. RESULTS: With the Microscan, the most accurate identification result was obtained after a modified inoculation method of the panel with a bacterial suspension prepared in 0.85% saline; the identification rate was 100%. The identification rate of Vitek II system was 96.7%; two strains of V. vulnificus were misidentified as V. harveyi and V. alginolyticus. CONCLUSIONS: These results indicate that both Microscan and Vitek II are adequate for the identification of clinical isolates of V. vulnificus, but for the identification by the Microscan a modified inoculation method should be used by suspending the organisms in 0.85% saline.