This study assessed the reliability of a color measurement method using images obtained from a charge-coupled device (CCD) camera and a stereoscopic loupe. Disc-shaped specimens were created using the composite Filtek Z350 XT (shades DA1, DA2, DA3, and DA4) (n = 3). CIELAB color coordinates of the specimens were measured using the spectrophotometer SP60 over white and black backgrounds. Images of the same specimens were taken using a CCD camera attached to a stereoscopic loupe. The color of the image was measured (red-green-blue [RGB]) using an image processing software and converted to CIELAB coordinates. For each color coordinate, data from images were adjusted using linear regressions predicting those values from SP60. The whiteness index for dentistry (WID) and translucency parameter (TP00) of the specimens as well as the color differences (ΔE00) among pairwise shades were calculated. Data were analyzed via repeated-measures analysis of variance and Tukey's post hoc test (α = 0.05). Images obtained using the loupe tended to be darker and redder than the actual color. Data adjustment resulted in similar WID, ΔE00, and TP00 values to those observed for the spectrophotometer. Differences were observed only for the WID of shade DA3 and ΔE00 for comparing DA1 and DA3 over the black background. However, these differences were not clinically relevant. The use of adjusted data from images taken using a stereoscopic loupe is considered a feasible method for color measurement.