In an assessment of the reproducibility of three-dimensional masticatory movements by an optoelectronic method, the tests conducted on the same individuals at different times produced results that were almost identical. However, some significant differences were found in the mandibular movement parameters when the subjects chewed different foods. This was mainly reflected in a longer opening time and higher values for the traversed 3D distance during mastication of chewing gum compared to peanuts. A movement analysis system was linked with a computer, and a program was developed for analysis of chewing parameters including mandibular velocity and dimensions. A reference system compensated for head movements. Every step in the recording procedure--calibration, collection, processing and presentation of the data--can be automatically controlled by the computer. The method proved to be precise and accurate in repeated recordings and apparently suitable for objective posttreatment and long-term evaluation of chewing performance.