The K121Q polymorphism of the glycoprotein PC-1 gene was recently reported to associate with insulin resistance (IR) in an all-Caucasian, Sicilian population. Given black-white differences in plasma insulin and IR, we compared the prevalence of the KK, KQ, and QQ genotypes and their associations with insulin and IR in 2 large, biracial pediatric samples: 1 hospital-based (n = 301, 137 blacks and 164 whites) and 1 school-based (n = 639, 344 blacks and 295 whites). The Q allele frequencies in the hospital-based and school-based cohorts in black children were 0.80 and 0.77 and in the white children, 0.15 and 0.13. The K allele frequencies in the hospital-based and school-based cohorts in black children were 0.20 and 0.23 and in the white children, 0.85 and 0.87. Differences in allelic frequencies were highly significant (chi square test, P <.0001) for both the hospital-based cohort and the school-based cohort. Both cohorts were in Hardy-Weinberg equilibrium. Within race, after covariance adjusting for age and body mass index (BMI), there were no significant differences (P >/=.10) among the 3 PC-1 genotypes for insulin, glucose, or homeostasis model assessment (HOMA) IR. After covariance adjusting for age and BMI, black girls had higher insulin (P =.0007) and higher HOMA IR (P =.0002) than white girls. The K121Q polymorphism was not associated with insulin, glucose, or HOMA IR measures in black or white children. However, the QQ genotype was population-specific, encompassing most black children versus 1% to 3% of white children. As such, K121Q genotyping should be useful in epidemiology, population genetics, and forensic anthropology.