Purpose: The aim of this study was to evaluate the sensitivity of ultrasonography, integrating standard ultrasound and arthrosonography after injecting a saline solution into the glenohumeral cavity in cases of suspected rotator cuff tears., Materials and Methods: We prospectively examined 40 patients awaiting shoulder arthroscopy for suspected or diagnosed tears of the rotator cuff. A radiologist, unaware of the pre-operative diagnosis, performed an ultrasound scan on all the patients before and after the injection of saline solution into the glenohumeral cavity. The parameters considered were presence or absence of a rotator cuff injury; type of injury according to Snyder and its extent along the longitudinal and transverse planes; presence or absence of effusion into the articular cavity; subacromial/subdeltoid bursal distension. All the patients underwent arthroscopy either the same day or the day after the ultrasound examination., Results: Standard sonography showed 26 complete rotator cuff tears (type C according to Snyder), 2 partial tears (type B according to Snyder) and 12 intact rotator cuffs. Arthrosonography detected 31 complete rotator cuff tears (type C according to Snyder), 1 partial tear (type B according to Snyder) and 8 intact rotator cuffs.Arthroscopy identified 32 complete rotator cuff tears (type C according to Snyder), 1 partial tear (type B according to Snyder) and 8 intact rotator cuffs. Analysis of the results shows that, taking arthroscopy as the gold standard, the sensitivity of normal sonography is 81.2%, whereas that of arthrosonography is 96.8% (p < 0.05)., Conclusions: On the basis of the data obtained in this study, standard sonography, integrated with the injection of a saline solution into the glenohumeral cavity, considerably increases the diagnostic sensitivity for rotator cuff tears. The authors suggest that arthrosonography can be used in the event of suspected rotator cuff tears, when MRI is contraindicated.