Genetic variants in ankyrin 3 (ANK3) have recently been shown to be associated with bipolar disorder (BD). We genotyped three ANK3 SNPs previously found to be associated with BD (rs10994336, rs1938526, and rs9804190) in a Scandinavian BD case–control sample (N¿=¿854/2,614). Due to evidence of genetic overlap between BD and schizophrenia (SZ), we also genotyped these three SNPs in a Scandinavian SZ case–control sample (N¿=¿1,073/2,919). Combining our Scandinavian samples with an Icelandic sample (N¿=¿435 BD cases, 651 SZ cases, and 11,491 healthy controls), we found rs10994336 and rs9804190 to be nominally significantly associated with BD in this combined Nordic BD sample (N¿=¿1,289/14,105). Nominal P was 0.015/0.018 (fixed/random effect) for rs10994336 (Bonferroni corrected P¿=¿0.044/0.053) and 0.023 for rs9804190 (Bonferroni corrected P¿=¿0.069). None of the SNPs were significantly associated with SZ in the combined Nordic SZ case–control sample (N¿=¿1,724/14,410). These results further support that ANK3 is a susceptibility gene specific to BD and that more than one risk locus is involved. © 2011 Wiley Periodicals, Inc.