Biochemical evidence that pyruvate decarboxylase in S. cerevisiae might be constituted from two independently encoded subunits led us to question genetic evidence for a single structural gene. The main evidence for this was that three “structural” mutations appeared to be alleles of the same gene, PDC1 (Schmitt and Zimmermann 1982). We report that one of these mutations (pdcl-30) is not allelic either to other pdc1 alleles or to pdc2 mutations and therefore is has been renamed pdc3-30 thus identifying a new gene, PDC3. We have cloned the PDC3 gene, it represents a unique sequence in the genome and targeted integration shows tight linkage to the PDC3 locus. However, the size, abundance and regulation of the PDC3 transcript suggest that it does not encode a second structural gene. Possible functions for the PDC3 gene product are discussed.