The UCP2-UCP3 gene cluster maps to chromosome 11q13 in humans, and polymorphisms in these genes may contribute to obesity through effects on energy metabolism. DNA sequencing of UCP2 and UCP3 revealed three polymorphisms informative for association studies: an Ala-->Val substitution in exon 4 of UCP2, a 45 bp insertion/deletion in the 3'-untranslated region of exon 8 of UCP2 and a C-->T silent polymorphism in exon 3 of UCP3. Initially, 82 young (mean age = 30 +/- 7 years), unrelated, full-blooded, non-diabetic Pima Indians were typed for these polymorphisms by direct sequencing. The three sites were in linkage disequilibrium ( P < 0.00001). The UCP2 variants were associated with metabolic rate during sleep (exon 4, P = 0.007; exon 8, P = 0.016) and over 24 h (exon 8, P = 0.038). Heterozygotes for UCP2 variants had higher metabolic rates than homozygotes. The UCP3 variant was not significantly associated with metabolic rate or obesity. In a further 790 full-blooded Pima Indians, there was no significant association between the insertion/deletion polymorphism and body mass index (BMI). However, when only individuals >45 years of age were considered, heterozygotes (subjects with the highest sleeping metabolic rate) had the lowest BMI (P = 0.04). The location of the insertion/deletion polymorphism suggested a role in mRNA stability; however, it appeared to have no effect on skeletal muscle UCP2 mRNA levels in a subset of 23 randomly chosen Pima Indians. In conclusion, these results suggest a contribution from UCP2 (or UCP3) to variation in metabolic rate in young Pima Indians which may contribute to overall body fat content later in life.