A total occlusal convergence (TOC) angle (ϕ) of 6 degrees is a typical crown preparation requirement. It has been shown that this is difficult to achieve clinically. The present study therefore aims to compare the ability of students to judge different steepnesses, including a -1-degree undercut of prepared canines and molars, under clinical conditions with different analog tools. The complete dentures of a patient were duplicated without denture teeth 16, 23, 33, and 46. For each of these gaps, six crown stumps were milled with ϕ/2 = -1, 3, 6, 9, 12, and 15 degrees, each of which was insertable using mini-magnets. A total of 144 dental students, 48 students each from the 1st, 6th, or 9th semester, estimated these angles intraorally with various aiding tools: In addition to basic dental instruments, they used a parallelometer mirror, an analog clock dial with a 6-degree visualization, and a scale of tooth stumps with ϕ/2 between -1 and 15 degrees. The widely demanded 3 degrees were hardly recognized, but assumed to be steeper or even undercut. In contrast, the -1-degree divergent stump walls were predominantly estimated as parallel-walled or slightly conical. With increasing taper, the stumps tended to be classified as steeper, ie, "better." The additional tools did not result in a general improvement of the estimation performance. Students from higher semesters did not achieve better results. The present authors question the objectivity of an exclusively visual assessment of crown stump taper. It appears that dental training should at least focus on avoiding undercuts as the minimal prerequisite for an accurate intraoral scanning procedure. Digital control of the preparation angle by an intraoral scan and immediate clinical implementation of these results could help to produce appropriate preparations.