Disruptions in circadian rhythms have been described in mood disorders (MD), but the involvement of genetic variation in genes pertaining to the molecular circadian machinery in the susceptibility to MD has not been conclusively determined. We examined 209 single-nucleotide polymorphisms (SNPs) covering 19 circadian genes (ADCYAP1, ARNTL, ARNTL2, BHLHB2, BHLHB3, CLOCK, CRY1, CRY2, CSNK1E, DBP, NPAS2, NR1D1, PER1, PER2, PER3, RORA, TIMELESS, VIP, and VIPR2) in a sample of 534 MD patients (335 with unipolar major mood depression (MDD) and 199 with bipolar disorder (BD)) and 440 community-based screened controls. Nominally, statistically significant associations were found in 15 circadian genes. The gene-wide test, corrected for the number of SNPs analyzed in each gene, identified significant associations in CRY1 (rs2287161), NPAS2 (rs11123857), and VIPR2 (rs885861) genes with the combined MD sample. In the MDD subsample, the same SNPs in CRY1 and NPAS2 of the combined sample remained associated, whereas in the BD subsample CLOCK (rs10462028) and VIP (rs17083008) were specifically associated. The association with an SNP located 3' near CRY1 gene in MDD remained statistically significant after permutation correction at experiment level (p=0.007). Significant additive effects were found between the SNPs that were statistically significant at the gene-wide level. We also found evidence of associations between two-marker haplotypes in CRY1 and NPAS2 genes and MD. Our data support the contribution of the circadian system to the genetic susceptibility to MD and suggest that different circadian genes may have specific effects on MD polarity.