Candida dubliniensis is a recently described opportunistic pathogen which shares many phenotypic characteristics with Candida albicans but which has been reported to rapidly acquire resistance to azole antifungal drugs. Therefore, differentiation of C. dubliniensis from C. albicans becomes important to better understand the clinical significance and epidemiologic role of C. dubliniensis in candidiasis. We compared phenotypic methods for the differentiation of C. dubliniensis from C. albicans (i.e. the ability to grow at elevated temperatures, colony color on CHROMagar Candida medium, and carbohydrate assimilation patterns) to amplify the results of a polymerase chain reaction (PCR) assay using universal fungal primers to the internal transcribed spacer 2 (ITS2) region of rDNA and species-specific DNA probes in an enzyme immunoassay format (PCR-EIA). DNA sequencing of the ITS1 rDNA region was also conducted. The C. dubliniensis ITS2 probe correctly identified all C. dubliniensis isolates without cross-reaction with any other Candida species tested (mean A(650 nm) +/- SE, C. dubliniensis probe with C. dubliniensis DNA, 0.372 +/- 0.01, n = 22; C. dubliniensis probe with other Candida species DNA, 0.001 +/- 0.02 n = 16, P < 0.001). All other Candida species tested (C. albicans, Candida glabrata, Candida krusei, Candida parapsilosis, and Candida tropicalis) were also correctly identified by the PCR-EIA without any detectable cross-reactions among species. Phenotypically, C. dubliniensis isolates demonstrated an increased sensitivity to heat compared to C. albicans isolates. At 42 degrees C, only 50% of C. dubliniensis isolates grew compared to 73% of C. albicans isolates and, at 45 degrees C, 91% of C. dubliniensis isolates failed to grow compared to 64% of C. albicans isolates. C. albicans was more likely to demonstrate a dark green or blue green colony color on CHROMagar Candida medium obtained from Becton Dickinson (i.e. 100% of C. albicans isolates were dark green or blue green versus 64% of C. dubliniensis isolates) whereas no difference in the percentage of C. albicans or C. dubliniensis isolates producing dark green or blue green colony color was detected using CHROMagar Candida medium from Hardy Diagnostics (82% for both species). The API 20C AUX carbohydrate assimilation system incorrectly identified C. dubliniensis as C. albicans in all but three cases: remaining isolates were misidentified as C. albicans/C. tropicalis, C. tropicalis/C. albicans, and Candida lusitaniae/C. albicans. In all, 82% of C. albicans isolates and 100% of C. dubliniensis isolates assimilated trehalose; the latter finding was opposite to that reported for C. dubliniensis in the API 20C AUX profile index. Xylose and alpha-methyl-D-glucoside assimilation, respectively, were negative for 100 and 95% of C. dubliniensis isolates and positive for 100 and 91% of C. albicans isolates, confirming earlier reports that assimilation results for xylose and alpha-methyl-D-glucoside may be helpful in the discrimination of these two species. However, conventional phenotypic species identification tests required days for completion, whereas the PCR-EIA could be completed in a matter of hours. In addition, identification of Candida species by ITS1 rDNA sequencing gave 100% correspondence to the results obtained by the PCR-EIA, confirming the specificity of the PCR-EIA method. These data indicate that although a combination of phenotypic methods may help differentiate C. dubliniensis from C. albicans to some extent, the PCR-EIA can provide a simple, rapid, and unequivocal identification of the most medically important Candida species in a single test.