OBJECTIVE To examine the associations between common variations in the IL6R gene and circulating interleukin (IL)-6 levels and diabetes risk. METHODS We determined 10 linkage disequilibrium (LD)-tagging single nucleotide polymorphisms (SNPs) (SNP1 to SNP10) for the IL6R gene in a nested case-control study of 672 diabetic and 1,058 healthy European Caucasian women (IL-6 levels were measured in a subgroup of 1,348 women). RESULTS In both control and diabetic patients, polymorphisms within an LD block spanning approximately 42 kb were significantly associated with plasma IL-6 levels. A missense variant SNP7 in exon 9 (rs8192284, Asp358Ala) showed the strongest association (P = 0.0005 in control and P = 0.004 in case subjects). The corresponding false-discovery rates, which accounts for multiple testing, were 0.008 and 0.02, respectively. We inferred five common haplotypes to capture 94% allele variance of the LD block using SNP5, -7, -8, -9, and -10. Compared with the most common haplotype 12111 (one codes the common and two codes the minor alleles), haplotypes 11211 [difference in log(IL-6) = -0.11 (95% CI -0.23 to -0.01); P = 0.01] and 21122 (-0.15 [-0.27 to -0.03]; P = 0.01) were associated with significantly lower IL-6 levels (global test, P = 0.01). However, IL6R genotypes were not significantly associated with the risk of type 2 diabetes. CONCLUSIONS IL6R genetic variations, especially SNP7 (rs8192284, Asp358Ala), were significantly associated with plasma IL-6 levels but not with diabetes risk in women. The strong associations between IL6R genetic variability and IL-6 concentrations deserve further investigation.