We noted that SNP rs5016282, the second best hit of our genome-wide association study (GWAS) [Hinney et al., 2011], is most likely wrongly located to the gene GRM5. The SNP thus needs to be removed from the paper.After our publication we detected that the genotypes of SNP rs5016282 could not be confirmed by Sanger re-sequencing. Reassessment of the probe used on the Illumina chip Quad 660 (personal communication) and our ARMS-PCR primers [for primers see Hinney et al., 2011] showed that the allele specific probe/primers can bind correctly to the intronic sequence near SNP rs5016282 in GRM5 (location: chr11: 88,381,258-88,381,307; hg19) and also to an independent region within a pseudo-gene for GRM5 (BC142657 alias LOC440040, chr11: 49,594,435-49,594,484, hg19). The sequences of GRM5 and the pseudo-gene are, for the region amplified by our ARMS-PCR, nearly identical, so that both, the Illumina chip and the ARMS-PCR detected the same two independent loci. Genotyping with both methods revealed nearly identical results [see: Hinney et al., 2011]. Of note: the SNP is not reported for the pseudo-gene. However, we assume that the assignment of SNP rs5016282 to the genomic region of GRM5 is incorrect. As SNP rs5016282 in dbSNP (http://www.ncbi.nlm.nih.gov/sites/ entrez?db=snp) was mainly analysed with Illumina GWAS chips, allele and genotype frequencies in dbSNP are most likely incorrect as well. We thus also submitted a correction to dbSNP. Article first published online in Wiley Online Library (wileyonlinelibrary.com)