The object of this study was to measure any mandibular flexure occurring in the horizontal plane, when R.A.P. recordings were made with an 'anterior jig' (Lucia, 1964), chin-point guidance and patient-applied muscle force. The rationale for the experimental method was that any arch-width change noted at the tooth level reflected a flexure of the mandible. Ten subjects participated and the mean lateral flexure of the mandible in the horizontal plane was 0.073 +/- 0.028 mm. As a corollary to the study, mean medial flexure of the mandible in wide opening movements was found to be 0.093 +/- 0.044 mm, which was consistent with earlier workers' results. On the basis of the results obtained, restorations constructed to muscle-R.A.P. recordings could conceivably present as occlusal interferences and, indeed, articulators may require modification so as to allow for mandibular resilience.