Introduction: The specific, natural color of the future prosthodontic restoration is of a great importance not only for the patient, but for the dentist too. Therefore, an accurate shade assessment is an important part of the protocol in everyday dental practice when using ceramic materials and composits. There are different ways of color evaluation in dentistry. Dentists mostly use visual tools and analyse the natural teeth color using different shadeguades, but this subjective method is difficult to perform because it is influenced by ambiental light, observer's characteristics, dentists' knowledge, experience, skills etc. VITA Classical A1-D4 and VITA 3D-Master shadeguides are the most frequently used visual method of tooth color assessment. Regardless of its widespread use, VITA Classical A1-D4 shade guide demonstrates many clinical difficulties with its colors not grouped methodically, and with the progress of 3D-Master shadeguide a more systematic approach was introduced in prosthodontics. Using this tool it is important to select tooth shade by determing three basic characteristics - hue, value and chroma under the standardized artificial daylight and color temperature of 5500 K. Aim of the study: The aim of this study was to use TTB (Tooth Training Box) device for the purpose of assessing tooth color according to VITA 3D Master shadeguide in four different dental occupational groups with different levels of clinical experience, with no prior knowledge of using this system. Materials and methods:A total of 120 subjects participated in the study, 30 in each of four groups as follows: prosthodontists (PR), prosthodontic residents (RE), dental technicians (DT) and dental students (ST). Subjects underwent three exercises assessing lightness, saturation and hue consecutively and a final test assessing all three color characteristics simultaneously, using TTB device in which lightness, chroma and hue needed to be assessed for each of 15 tested teeth. The device used the standard illumination and 52 teeth different in 3D Master sistem lightness, saturation and hue and was connected to a computer recording the results. The number of mistakes, their mistake ratio and time in exercises, the result of the final test, time and difference in color assessment were recorded according to gender, age and level of clinical experience. The influence of exercises on the results of the final test was also tested. Results:Female specialists rated lightness and saturation significantly worse in exercises and female students rated lightness, saturation and hue in exercises faster, were more accurate in the final test than students (p<0,05). Male specialists were more accurate in final test, made less mistakes and needed more time to assess lightness (p<0,05). Female students were more accurate than male students, achieved lower color differences and meeded less time to assess saturation than (p<0,05). Female dental technicians needed less time to assess lightness in final test (p<0,05). Specialists were more accurate in second exercise where they needed least time, but were the least accurate in third exercise and needed more time (p<0,05). Male prosthodontic residents were statistically less accurate and needed more time to finish first and second exercises while male students needed stiatistically less time to finish second exercise (p<0,05). The practice of the new protocol strongly influenced the final test result in the group of dental technicians and students, the time required to complete the final test in the group of specialists and students, and the range of mistakes in the final test in the group of dental technicians (p>0,05). According to the level of clinical experience SCORE values, time required to complete the final test, and differences in color assessment did not differ significantly (p> 0.05). Dental technicians made less mistakes, were the most accurate in assessing color and spent more time to assess it (p> 0.05). Students made more mistakes and solved the final test the fastest (p> 0.05). The lightest (1M1) and darkest colors (5M2, 5M3) were estimated the most accurately. The least accurate colors estimated in this study were 3M1, 3M2, 2M3 i 3L2.5. Conclusion:The results of this study demonstrated that color assessment practice strongly influences the final result and that clinical experience is not a significant factor in accuracy in color determination.