This paper presents preliminary results demonstrating the use of the sugarcane expressed sequence tag (EST) database (SUCEST) to detect single nucleotide polymorphisms (SNPs) inside 6-phosphogluconate dehydrogenase genes (Pgds). Sixty-four Pgd-related EST sequences were identified and partitioned into two clear-cut sets of 14 and 50 ESTs, probably corresponding to two genes, A and B, respectively. Alignment of A sequences allowed the detection of a single SNP while alignment of B sequences permitted the detection of 39 reliable SNPs, 27 of which in the coding sequence of the gene. Thirty-eight SNPs were binucleotidic and a single one was trinucleotidic. Nine insertions/deletions from one to 72 base pairs long were also detected in the noncoding 3' and 5' sequences. The soundness and the consequences of those preliminary observations on sequence polymorphism in sugarcane are discussed.