This study was performed to test the hypothesis that genetic variation in the promoter of the glucose transporter 2 (GLUT2) might predispose to prediabetic phenotypes or type 2 diabetes. A total of 1611 bp comprising the minimal promoter region of the GLUT2 gene were examined by combined single-strand conformational polymorphism and heteroduplex analysis followed by direct sequencing of identified variants on genomic DNA from 96 randomly recruited Danish type 2 diabetic patients. We identified 4 nucleotide variants, -447g-->a, -149c-->a, -122t-->c, and -44g-->a. None of the variants were positioned in known or presumed transcription factor binding sites, TATA-box, or transcriptional start site. Association studies of the -149c-->a, -122t-->c, and -44g-->a variants revealed that the variants were as prevalent in 320 type 2 diabetic patients [11.0% (95% confidence interval, 8.4-13.6), 9.8% (7.4-12.2), and 29.0% (24.4-33.6), respectively] as in 241 age-matched glucose-tolerant subjects [13.1% (9.8-16.4), 11.2% (8.3-14.1), and 33.4% (28.8-38.0), respectively]. The -447g-->a mutation was only identified in a single diabetic patient and did not show cosegregation with diabetes in the family of the proband. The three common variants showed in a primary genotype-phenotype study comprising 241 glucose-tolerant middle-aged subjects association to increased plasma glucose levels during an oral glucose tolerance test. However, this result could not be replicated in a second sample of 298 60-yr-old glucose-tolerant subjects. In conclusion, we found no evidence supporting the hypothesis that genetic variability in the minimal promoter of the GLUT2 is associated with type 2 diabetes or prediabetic phenotypes in the Danish population.