BACKGROUND Accuracy of techniques for recording the orientation of the maxillary arch to the articulation of the temporomandibular joints has been reported. The variability contributed by the dentist and the equipment within a technique may also contribute to technical error. OBJECTIVE This study investigated the variability of a group of dentists who used an arbitrary ear-face-bow instrument to mount a maxillary cast. METHODS The same maxillary cast on a single experimental subject was mounted on an articulator with a common arbitrary ear-face-bow instrument. This cast was equipped with 4 reference points for the measurement of changes of spatial orientation of the cast, which were determined through coordinate measurements with a machinist's microscope. A repeated measures analysis of variance was used to determine statistically significant changes in spatial orientation with a Wilks' Lambda test to compare mean values. RESULTS A confidence interval of 95% demonstrated that any dentist might expect a range of +/-1.2 mm error in using this instrument/articulator combination. The 3 dentists did not demonstrate any significant differences in the spatial orientation of their mountings in the vertical or horizontal directions. However, there were significant differences in their mountings in linear distance changes of the patient's posterior right side. CONCLUSIONS A range of inherent error attributable to the operator using this instrument was recorded.