Acceptable reliability is a prerequisite for inclusion of physical examination tests in clinical examinations of the painful shoulder. The aim of this study was to establish the intraexaminer and interexaminer reliability of measures of shoulder range of motion (ROM) and muscle force using a new hand-held dynamometer with the ability to standardize overpressure force during passive ROM tests. Forty consecutive subjects with shoulder pain were recruited, and tests were performed by two physiotherapists. Tests included active ROM elevation, passive ROM glenohumeral abduction and external rotation and resisted abduction and external rotation. All tests demonstrated high levels of intraexaminer reliability (ICC 0.85-0.99; LOA 6-24° and 1.1-7.0 kg). Highest levels of interexaminer reliability reliability were observed for measures of active ROM flexion (ICC 0.88-0.95; LOA 14-22°). Passive ROM tests demonstrated 'moderate - substantial' interexaminer reliability (ICC 0.45-0.62; LOA 25-34°). The ICCs for resisted tests ranged from 0.68 to 0.84, and LOA ranged from 3.2 to 8.5 kg. Active ROM flexion demonstrated high levels of both intra- and interexaminer reliability. Measures of passive ROM and peak isometric force demonstrated acceptable levels of intraexaminer reliability.