Polymorphisms in intron III of barley endosperm-specific b-amylase (Bmy1) have been associated with b-amylase activity and thermostability and are thought to have potential as a selective marker for breeding elite malting cultivars. Bmy1 intron III, from 40 barley genotypes, was sequenced. Four alleles were identified based upon insertion/deletions (indels) of 126-bp, 38-bp, 11-bp, and 21-bp. The Bmy1.a allele has the 126-bp, 38-bp, and 21-bp indels. Bmy1.b only has the 38-bp indel. Bmy1.c has the 38-bp, 11-bp, and 21-bp indels. Bmy1.d, found in only one genotype, is missing the 38-bp indel. Thirty-nine genotypes were assayed for b-amylase activity and thermostability (% remaining after exposure to 60°C for 10 minutes). Genotypes with the Bmy1.a, Bmy1.b, Bmy1.c, and Bmy1.d alleles had averages of 26.7, 23.8, 38.5, and 56% residual activity, respectively. Each allele had statistically different thermostabilities according to the Kruskal-Wallis test. Genotypes with Bmy1.a, Bmy1.b, Bmy1.c, Bmy1.d alleles had averages of 1181, 1460, 1058, and 2739 U/g flour, respectively. Each allele had statistically different activities using the Kruskal-Wallis test, except between Bmy1.a and Bmy1.c. These data support the use of intron III as a selective tool. However, when the data within each intron allele are analyzed, the conclusions are different. Using the least significant difference (LSD) test on the activities and thermostabilities within each allele (i.e. Bmy1.a, Bmy1.b, and Bmy.c), statistically significant differences were found. Therefore, we conclude that the third intron is not always a reliable marker for use in selecting elite malting cultivars.