Deficiencies in the human visual percep-tion system have challenged the efficiency of the visual shade-matching protocol. The aim of this study was to evaluate the repeatability and reliability of human eye in visual shade selection. Fifty-four volunteering dentists were asked to match the shade of an upper right central incisor tooth of a single subject. The Vita 3D-Master shade guide was used for the protocol. Before each shade-matching procedure, the definitive codes of the shade tabs were hidden by an opaque strip and the shade tabs were placed into the guide randomly. The procedure was repeated 1 month later to ensure that visual memory did not affect the results. The L*, a* and b* values of the shade tabs were measured with a dental spectrophotometer (Vita Easyshade) to produce quantitative values to evaluate the protocol. The paired samples t-test and Pearson correlation test were used to compare the 1st and 2nd selections. The Yates-corrected chi-square test was use to compare qualitative values. Statistical significance was accepted at P < 0·05. Comparing baseline and 1st month records, statistical significance (P < 0·001) was found among qualitative data regarding repeatability on a yes/no (1/0) basis, revealing a very low percentage of repeatability (11·1%). Comparing baseline and 1st month records, statistical significance was not found (P = 0·000) among the L*, a*, b* and ΔE variables. These results indicate that dentists perform insufficiently regarding repeatability in visual shade matching, but they are able to select clinically acceptable shades.