The purposes of this study were (1) to examine intratester and intertester reliability of measurement of anterior and posterior shoulder tightness in patients with stiff shoulders (SS), and (2) to assess construct validity by determining the relations between shoulder tightness, shoulder range of motion (ROM), and self-report measures of functional limitation. Anterior and posterior shoulder tightness were measured by two testers in below-chest abduction and cross-chest adduction tests with an inclinometer, respectively, in 16 patients with SS. Both the intratester and intertester reliability for shoulder tightness measurements were good (intratester ICC=0.84 and 0.91; intertester ICC=0.82 and 0.89). The limit of intra-tester and inter-tester agreement (mean, -0.3+/-4.4 degrees ) was acceptable as compared to the standard deviations of the measurements (range, 6.2-7.4 degrees ). Significant relationships between internal rotation and posterior shoulder tightness (R2=0.448, P=0.002), external rotation and anterior shoulder tightness (R2=0.499, P=0.003), and functional disabilities and posterior shoulder tightness (R2=0.432, P=0.006) were found. Significant correlations between shoulder internal rotation and cross-chest adduction, shoulder external rotation and below-chest abduction were observed, indicating that internal and external rotations might be related to posterior and anterior shoulder stiffness. The study also revealed significant relationship between functional disabilities and cross-chest adduction. Below-chest abduction and cross-chest adduction were found to provide reliable data. The construct validity of the abduction and adduction tests is supported by the relationship among these measurements, shoulder ROM, and functional disabilities in patients with SS.