Vitamin D, a pleiotropic hormone essential for calcium homeostasis, has generated widespread interest due to associations with numerous health outcomes. Cross-sectional studies of vitamin D and lung function reported strong, positive associations, but representative, longitudinal population-based studies are lacking, and biological mechanisms are unclear. Lung function decline is the primary characteristic of chronic obstructive pulmonary disease (COPD), the 3rd leading cause of mortality in the United States; given limited treatments to delay progression, identifying preventative approaches is critical. This work aims to elucidate determinants of vitamin D status, and investigate the role of vitamin D as a determinant of lung function. First, we explored genetic and non-genetic determinants of serum vitamin D [25(OH)D] in African Americans. Approximately 25% of 25(OH)D variability was explained by non-genetic factors, and multivitamin supplement use was the strongest predictor. A single nucleotide polymorphism (SNP) in the vitamin D binding protein modified the effect of multivitamin supplement use on 25(OH)D. About 23% of 25(OH)D variability was estimated to be attributable to genetic variation, with replication in a separate cohort. However, the influence of genetic ancestry made an exact estimate impossible; further exploration of genetic determinants of 25(OH)D in African Americans is needed. Second, potential mechanisms for vitamin D-lung health associations were explored through a cross-sectional study of SNPs in 13 candidate vitamin Dresponsive genes. SNPs in SGPP2, a phosphatase in the sphingosine-1-phosphate signaling pathway, were associated with lung function and COPD risk. Further, we identified an association between SNPs in SGPP2 and lung-tissue specific expression of SGPP2. While specific mechanisms remain to be investigated, SGPP2 is a promising vitamin D-responsive candidate gene. Finally, associations between variants in vitamin D metabolic genes, serum 25(OH)D and lung function were explored in the Framingham Heart Study. SNPs in four vitamin D metabolic genes were associated with rate of change in FEV1, but there was no association between 25(OH)D and rate of change in FEV1 in the Third Generation cohort, a group of largely vitamin D sufficient middleaged adults. Future studies should consider the influence of baseline nutritional status and underlying genetic variation on vitamin D-disease associations.