We compared the recovery of yeasts from clinical specimens cultured on routine bacteriological media to the recovery of yeast from specimens cultured on a selective fungal medium (Sabouraud agar). The use of Sabouraud agar was especially important in cases of mixed cultures, since in such cases yeast was recovered on bacteriological media from only 50% of 44 yeast-positive pus specimens and from 22.5% of 22 yeast-positive throat specimens. The use of a selective fungal medium is therefore necessary to ensure the detection of yeast in specimens containing a mixture of bacteria and yeasts. As a result, clinicians must request yeast isolation when clinically indicated, and the microbiological laboratory must add a selective fungal medium when clinically significant yeasts are likely to be encountered. It is also important that selective fungal media be used in clinical studies of yeast infections.