This clinical outcomes study aimed to determine the commonly used shoulder tests predictive for rotator cuff tear (RCT). The study was carried out in 2 parts involving 400 patients who attended the senior author’s practice (n=100/group). Part I compared patients with isolated RCTs and a “No-RCT” group. Part II compared RCT patients (some with isolated RCTs, others with RCTs and combined pathology) and No-RCT patients. Patients were imaged, preoperatively examined with clinical shoulder tests, and given presurgical diagnostic arthroscopy if warranted by their shoulder injury. Part I revealed 4 clinically diagnosed features significantly more common in patients with isolated RCTs than in No-RCT patients. Part II validated the findings, showing remarkably similar results. The drop arm sign (10% sensitivity, 100% specificity) affected a small proportion of subjects but where present was a strong indicator for RCT. Three other diagnostic features [supraspinatus weakness, weakness in external rotation, and impingement (SERI) tests] were highly predictive of RCT when all were positive in a subject (likelihood ratio=12:1). If all SERI tests were negative, RCT was unlikely to be present (likelihood ratio=0.06:1). The predictive power of this clinical test combination compares favorably with predictive powers reported for magnetic resonance imaging and ultrasound.