Abstract Introduction The prevalence of vitamin D deficiency (VitD) in the United States is 41 percent, with the highest rate among Blacks 82 percent. Vitamin D deficiency has been linked to chronic diseases. The extent to which the association between light exposure and vitamin D serum levels can vary by individual’s race/ethnicity of which has not been studied at a national level. We aim to explore the associations of ambient light exposure between race/ethnicity and vitamin D. Methods The study used data from the National Health and Nutrition Examination Survey (2013-14). For detection of 25-hydroxyvitamin D3 and 25-hydroxyvitamin D2 nmol/L, ultra-high-performance liquid chromatography-tandem mass spectrometry was performed based on serum samples from adults aged ≥ 18 years. Light levels (lux) data were gathered using 24-hour actigraphic monitoring over /a 7day period. Weighted generalized linear models were fitted examining association between light exposure and VitD adjusting for age, sex, family income/poverty ratio, education, employment, house tenure, marital status, birthplace, number of people in household, smoking, physical activity, and sedentarity. To compare this association across race/ethnicity, a product term between lux and race/ethnicity was included in adjusted models. Results Among 4,251 participants, White adults had the highest levels of VitD (mean=76.0; se=1.3), then other/multiple races (mean=65.1; se=2.2), Asians (mean=62.5; se=1.4); Hispanics (mean=57.4 nmol/L; se=1.6), and Blacks (mean=50.1; se=1.4). Regression analysis revealed estimated mean VitD of 64.9 nmol/L and positive association between light exposure and VitD ( 0.020). Blacks had significantly lower VitD levels ( -19.3) followed by Asians (-12.1) and Hispanics (-12.6) (all p-values <0.001). The association between light exposure and VitD depended on participant’s race/ethnicity Conclusion To our knowledge, this is the first study showing associations between objectively measured light exposure and VitD serum levels using a large representative sample of the US population. Although the study revealed racial/ethnic disparities in VitD levels, light exposure was associated with VitD even when race/ethnicity was adjusted for in the model. Further research on racial/ethnic differences in VitD is warranted. Support (If Any) R01HL142066, R01HL095799, RO1MD004113, R01HL152453, R25HL105444