Analysis of amylase RNA levels in the anterior and posterior midgut regions of flies from the Amy1,6 mapA and c Amy2,3 mapC strains of D. melanogaster, reared on yeast and on yeast supplemented with glucose, indicates that the trans-acting map gene controls the abundance of amylase RNA tissue-specifically, i.e., in the adult posterior midgut. This is consistent with the view that its role in controlling Amy expression is that of a transcription factor. Dietary glucose represses Amy expression in the anterior and posterior midgut regions of adults, reducing the abundance of amylase RNA, which suggests that it also controls Amy transcriptional activity. However, the mechanism for glucose repression appears to act systemically in the midgut, in a manner that is independent of the effects of map on Amy expression. A new glucose repressible TU was identified that is located just proximal to the Amy locus in region 54A of polytene chromosome 2R. It is transcribed in the direction opposite to that of the proximal Amy gene and encodes an RNA about 1500 bases long. Its RNA is expressed in both larvae and adults of the above strains of D. melanogaster, but the nature of the product it encodes is unknown. We speculate that all three genes in the cluster at 54A, namely the two Amy gene copies and the new glucose repressible TU, are coordinately controlled by the same mechanism that regulates Amy gene expression in response to dietary glucose. Somatic transformation experiments suggest that 5' cis-regulatory mechanisms required for the correct spatial expression of the proximal and distal Amylase genes from a Canton-S strain of D. melanogaster, Amy-p1 and Amy-d3, are located within 450 bp and 463 bp of their respective translation start sites. These regions also contain sequences responsive to dietary glucose repression, which is mediated at the DNA level of exogenous Amy genes in somatically transformed larvae reared on a yeast + glucose diet. A positive activator is located in the upstream region of Amy-d3 between the nucleotide pairs at -365 and -252 from the translation start site, but a comparable activator does not appear to exist in the upstream region of Amy-p1. Deletion analysis of the 5' sequence flanking the coding region of Amy-d3 indicates 125 nucleotide pairs of flanking DNA is sufficient for its functional activity. A model is proposed for coordinate control, in part, of the duplicated Amy genes.