Erythema (i.e. visible redness) and DNA damage caused by ultraviolet radiation (UVR) in human skin have similar action spectra and show good correlation after a single exposure to UVR. We explored the potential to use instrumental assessments of erythema as a surrogate for DNA damage after repeated exposures to UVR. We exposed 40 human subjects to three different exposure schedules using two different UVR sources. Cyclobutane-pyrimidine dimers (CPDs) in skin biopsies were measured by immunofluorescence, and erythema was assessed by both the Erythemal Index (EI) and the Oxy-hemoglobin (Oxy-Hb) content. Surprisingly, the skin with the highest cumulative dose ended up with the lowest level of DNA damage, and with the least erythema, as assessed by Oxy-Hb (but not EI) 24 h after the last UV exposure. Although the level of CPDs, on average, paralleled Oxy-Hb (R² = 0.80–0.94, P = 0.03–0.11), the correlation did not hold for the pooled individual measurements (R² = 0.009, P = 0.37) due to potential individual differences in UV-induced photoadaptation. We suggest that the methodology may be optimized to improve the correlation between DNA damage level and erythema to enable noninvasive risk assessment based on erythema/Oxy-Hb content for individual human subjects.