Fructosamine and glycated albumin are potentially useful alternatives to hemoglobin A1c (HbA1c) as diabetes biomarkers. The genetic determinants of fructosamine and glycated albumin, however, are unknown. We performed genome-wide association studies of fructosamine and glycated albumin among 2,104 black and 7,647 white participants without diabetes in the Atherosclerosis Risk in Communities (ARIC) Study and replicated findings in the Coronary Artery Risk Development in Young Adults (CARDIA) study. Among whites, rs34459162, a novel missense single nucleotide polymorphism (SNP) in RCN3, was associated with fructosamine (P = 5.3 × 10−9) and rs1260236, a known diabetes-related missense mutation in GCKR, was associated with percent glycated albumin (P = 5.9 × 10−9) and replicated in CARDIA. We also found two novel associations among blacks: an intergenic SNP, rs2438321, associated with fructosamine (P = 6.2 × 10−9), and an intronic variant in PRKCA, rs59443763, associated with percent glycated albumin (P = 4.1 × 10−9), but these results did not replicate. Few established fasting glucose or HbA1c SNPs were also associated with fructosamine or glycated albumin. Overall, we found genetic variants associated with the glycemic information captured by fructosamine and glycated albumin as well as with their nonglycemic component. This highlights the importance of examining the genetics of hyperglycemia biomarkers to understand the information they capture, including potential glucose-independent factors.