Accurate assessment of head motion can be a useful tool in clinical studies. Since the head moves on a combination of axes in the cervical spine, evaluation of neck motion is difficult. Assessment of cervical mobility is further complicated because of inadequate reference points on the head from which to measure. Hence, numerous methods for approximating cervical range have been devised. These methods include visual estimation, radiographic analysis, schematography, photography, and a variety of goniometer devices. Disadvantages of these techniques are lack of accuracy and objectivity, radiation exposure, expense, time consumption, and equipment availability. To measure cervical mobility, a standard gravity goniometer with spirit level and head adapter was used, which allowed stabilization. The gravity goniometer can be obtained in a variety of sizes at most hardware stores. The head adapter consists of a wood block into which an arc is carved and elastic straps suspended for securing on the head. The reliability of this instrument was tested and compared to the universal goniometer, and correlation coefficients were determined. When two experienced examiners used the universal goniometer to assess cervical motion, significant intraclass correlation coefficients (ICC) were found with three of the six criteria measures (p less than 0.05). When one experienced and one novice examiner used the gravity goniometer with head adapter, highly significant ICC values were found for all six criteria measures (p less than 0.01). A single experienced examiner comparing both instruments on the same subjects produced significant ICC values in four of the six criterion measures (p less than 0.01).(ABSTRACT TRUNCATED AT 250 WORDS)