Abstract Background: Disruption of sleep duration or circadian rhythms has been associated with an increased risk of prostate cancer. Studies have suggested that circadian genes may be involved in regulating cancer-related pathways, such as cell proliferation, DNA damage response and apoptosis. As such, we hypothesized that genetic susceptibility to prostate cancer may be in part due to common variation in the core circadian rhythm genes. We assessed the association of genetic variation in core circadian rhythm genes with risk of prostate cancer and with urinary levels of melatonin among men in the prospective Age, Gene/Environment Susceptibility (AGES)-Reykjavik study. Methods: The study population was 1,352 men in the AGES-Reykjavik cohort with data from a previous genome wide association study. We examined 96 single-nucleotide polymorphisms (SNPs) across twelve genes chosen to represent common variation in genes that control the circadian rhythm (CLOCK, neuronal PAS domain protein 2 (NPAS2), aryl hydrocarbon receptor nuclear translocator-like (ARNTL), cryptochrome 1(CRY1), cryptochrome 2 (CRY2), period 1 (PER1), period 2 (PER2), period 3 (PER3), casein kinase 1-epsilon (CSNK1E)), two melatonin receptor genes (MTNR1A and MTNR1B), and the Timeless (TIM) gene. Melatonin levels were determined from first morning void urine samples using an ELISA assay. Prostate cancer (N = 138) was identified from hospital records using ICD codes; death from prostate cancer (N = 24) was determined from nationwide death records. We used unconditional logistic regression to evaluate the association between each SNP and overall and lethal prostate cancer. Linear regression was used to assess the association between each SNP and log-transformed melatonin levels. Finally, to test whether multimarker SNP-sets in the pathway were associated with prostate cancer or melatonin levels, we used a kernel machine test. Results: None of the 96 SNPs evaluated were significantly associated with prostate cancer overall. The variant allele in two SNPs in CRY1 (rs7297614 and rs1921126) and one in PER1 (rs2289591) were nominally (p-value <0.05 without adjusting for multiple testing) associated with an elevated risk of lethal prostate cancer, while SNPs in NPAS2 (rs3754674), ARNTL (rs969485) and PER2 (rs10462023) were nominally associated with a reduced risk of lethal disease. Among men without cancer, 4 SNPs in TIM and 6 in NPAS2 were nominally associated with lower levels of urinary melatonin; 2 SNPs in PER3 and 1 in CSNK1E were associated with higher levels of urinary melatonin. Pathway analyses showed that variability across the individual genes CRY1 and PER2 were nominally associated with lethal prostate cancer (p-value 0.01 and 0.05, respectively). For melatonin levels, the set of SNPs across all 12 genes and TIM was nominally associated with urinary melatonin levels (p-value 0.05 and 0.02, respectively). Conclusion: Variation in circadian genes, such as PER1, CRY1, and CRY2, has been previously associated with increased prostate cancer risk; variation in NPAS2 has been associated with reduced prostate cancer risk. Similar to previously reported literature, we observed several suggestive associations between variation in genes that regulate the circadian rhythm and lethal prostate cancer risk as well as urinary melatonin levels. Our findings provide further support for a role of circadian disruption in prostate cancer progression. Citation Format: Sarah Coseo Markt, Unnur A. Valdimarsdottir, Lara G. Sigurdardottir, Irene M. Shui, Jennifer R. Rider, Julie L. Kasperzyk, Steven W. Lockley, Charles A. Czeisler, Meir J. Stampfer, Thor Aspelund, Albert Vernon Smith, Vilmundur Gudnason, Lorelei A. Mucci. Variation in circadian rhythm genes and risk of prostate cancer in the AGES-Reykjavik cohort. [abstract]. In: Proceedings of the Eleventh Annual AACR International Conference on Frontiers in Cancer Prevention Research; 2012 Oct 16-19; Anaheim, CA. Philadelphia (PA): AACR; Cancer Prev Res 2012;5(11 Suppl):Abstract nr B90.