BACKGROUND: Carotenoid levels measured in the skin and retina are increasingly being used as biomarkers of general health, function, and disease. Whether they are directly interchangeable, however, is not clear. Several previous studies comparing the measures in adults and children have found inconsistent results. OBJECTIVE: The objective of the present observational study is to assess the relation between carotenoid levels measured in the skin and retina. METHODS: We assessed the relation between the two measurements over an extended time period and multiple testing sessions. 21 healthy participants were assessed (mean age = 27.0, SD = 9.8). Macular carotenoids were measured using a standard method based on customized heterochromatic flicker photometry. Skin carotenoids were measured using pressure-mediated reflection spectroscopy. RESULTS: Both the retinal and skin measures showed high reliability (intraclass correlations >
0.90) and stability over time. The two measures, when averaged over 5 testing sessions, were moderately but significantly correlated (r = 0.55, p = 0.01). CONCLUSIONS: Skin carotenoid levels accounted for approximately a third of the variance in retinal carotenoid density. Carotenoids in both skin and retina likely share many factors that both promote and interfere with accumulation. Each, however, also likely has some unique predictive characteristics as a biomarker. Therefore, when using physiological carotenoids as a biomarker, the outcome of interest should determine whether to measure carotenoids dermally or in the retina.