Background: The objective of this study is to examine the association between special tests for subscapularis tears and both pain and tear severity. Methods: Nine hundred and two patients (145 with isolated lesions and 757 with combined lesions) diagnosed with arthroscopically confirmed subscapularis tears between 2010 and 2022. Subscapularis tear severity was classified according to Lafosse classification type I, II, III and IV. Preoperative Visual Analogue Scale (VAS) pain score at rest of the patients, flexion and abduction range of motion (ROM) of the shoulder and the results of the Lift-off test (LOT), Belly Press test (BPT), Bear Hug test (BHT), and Empty Can test (ECT) were documented. Results: The sensitivity of LOT in isolated and combined tears was 70.3 % and 69.5 %, 45.6 % and 41.2 % of BPT, 72.1 % and 72.6 % of BHT, 81.2 % and 84.1 % of ECT, respectively. There was no association between the severity of the subscapularis tear and the sensitivity in all special tests (p > 0.05). Only the ECT was observed to be associated with pain in the multivariate logistic regression analysis (OR = 33.1, p < 0.001). Conclusions: The special tests used to evaluate subscapularis tears are neither sensitive to the severity of the tear nor to pain except for ECT. BHT is the most successful test to detect any subscapularis tear in both isolated and combined tears. Pain severity was related to the presence of a rotator cuff lesion accompanying subscapularis lesion, but not with the severity of subscapularis or remaining rotator cuff tear. Study design: Levels of Evidence III, cross-sectional archive study. Clinical trial registration number: Since the study was a retrospective archive study, there was no clinical trial registration.