BACKGROUND: Face scanners provide a viable method for capturing a patient's face geometry. To optimize their accuracy, influencing factors, like the ambient light, need to be examined. METHODS: A human head model with eight pins attached to its surface was used to investigate the accuracy of four face scanning methods (Face Hunter, iPad, Medit i700, single camera photogrammetry) under three illumination levels (500 lx, 5000 lx, 20 000 lx). An industrial CT scan was used as reference. Two alignment-areas - full face (AL-FF) and spheres (AL-KG) and two investigation areas - center face (UB-CF) and full face (UB-FF), were used during the examination. The root-mean-square-error (RMSE) was employed as a measure. Separated by trueness and precision, a one-way ANOVA was performed with post hoc Games-Howell tests for each scanning method. RESULTS: All scanners showed significant differences between the illumination levels. For most test groups, the Face Hunter acquired its lowest RMSE values under 500 lx. The same can be said for the Medit i700, even though for trueness, differences to 5000 lx were random. Single camera photogrammetry performed better at higher illumination levels, but only random differences between 5000 lx and 20 000 lx were seen. For the iPad, different results for optimal illumination were found regarding trueness and precision, as well as the investigation areas. All accuracy results were labelled as highly reliable, except for the iPad´s trueness results. CONCLUSION: Scanner-dependent influence of ambient light was shown in this in-vitro study. Face Hunter and Medit i700 performed better under a darker illumination of 500 lx, whereas single camera photogrammetry needed brighter lighting. For the iPad no tested lighting situation showed clear advantages.