We developed a standardized DNA sequence-based approach for the accurate and timely identification of medically important fungi by sequencing polymerase chain reaction (PCR) products with a rapid automated capillary electrophoresis system. A simple DNA extraction method and PCR amplification using universal fungal primers was used to amplify ribosomal DNA from a range of clinical isolates and reference strains. The entire internal transcribed spacer (ITS) 1-5.8S-ITS2 ribosomal DNA region was sequenced using automated dye termination sequencing for 89 clinical isolates. These had previously been identified by traditional methods and included 12 ascomycetous yeast species, three basidiomycetous yeast species, eight dermatophyte species and two thermally dimorphic fungi, Scedosporium prolificans and S. apiospermum. Furthermore, 21 reference strains representing 19 different Candida species, Geotrichum candidum and Malassezia furfur were also sequenced as part of this study and were used either as standards for sequence-based comparisons, or as assay controls. Sequence-based identification was compared to traditional identification in a blinded manner. Of the clinical isolates tested, 88/89 had DNA sequences that were highly homologous to those of reference strains accessioned in GenBank, and 87/89 gave a sequence-based identification result that correlated with the traditional identification. In contrast to relatively slow conventional methods of identification, a sequence-based identification from a pure culture can be obtained within 24 h of a DNA extraction carried out after a minimal period of culture growth. We conclude that this approach is rapid, and may be a more accurate cost-effective alternative than most phenotypic methods for identification of many medically important fungi frequently encountered in a routine diagnostic microbiology laboratory.