The variability in the assessment of occlusal variables was investigated in eight subjects by the repeated registrations of four observers. The following variables were investigated in three different ways: sliding between the RCP (retruded contact position) and the ICP (inter-cuspal contact position) in the sagittal, vertical and lateral plane. Interferences during lateral movements on the working and balancing sides as well as interferences causing deviation of the mandible during protrusive movements were recorded. The intra- and inter-observer error for the measurement of sagittal and vertical distance RCP-ICP was within acceptable limits. There was no significant difference between the three methods used. The intra- and inter-observer agreement between duplicate recordings of the lateral distance RCP-ICP was high and there were no differences between the different measurement modalities. The inter-observer agreement was lower than the intra-observer agreement concerning presence/absence of balancing side interferences, and positive inter-observer agreement was only found on lateral movement more than 3 mm from the RCP or the ICP. The observer error for the variable working side interferences should be acceptable for future research, while the observer error for measurement of lateral deviation of the mandible upon protrusion was somewhat larger.