The purposes of this study were to determine the reliability of strength and self-reporting measures, the relationship of different strength measures to function, and the impact of rotator cuff pathology on patients' quality of life. Patients with nonoperated rotator cuff pathology (n = 36) and unaffected control subjects (n = 48) were assessed by use of the LIDO dynamometer to determine isometric and isokinetic (concentric and eccentric) strength of the shoulder rotators. The Shoulder Pain and Disability Index and Short Form-36 were self-reported by patients. Intraclass correlation coefficients (ICCs) were used to assess reliability, and Pearson correlations and multiple linear regression were used to determine the relationship between strength and function. The findings of this study include the following: (1) measures of self-reported physical disability had high reliability (ICC = 0.89); (2) the LIDO dynamometer reliably measured internal and external shoulder rotation strength in both concentric and isometric modes of testing (ICC = 0.78-0.94), whereas eccentric muscle actions had lower reliability; (3) all shoulder rotation strength measures were predictive of disability, with isometric external rotation strength being the most predictive (r = 0.56); and (4) the presence of rotator cuff pathology was highly predictive of impaired physical health quality of life (R(2) = 0.71, P < .001).