The effects of posterior plications associated with anterior shoulder instability surgery are still unclear both on shoulder range of motion (ROM) and on recurrence rate. The objective of this randomized study is to evaluate the influence of posterior-inferior plications, performed in association with repair of anterior Bankart lesion, on gleno-humeral (GH) range of motion. In a 24-month period, 40 patients were prospectively enrolled in this study. The criteria for inclusion were age between 17 and 40 years, traumatic unidirectional instability, no previous shoulder surgery, no more than three episodes of dislocation, no relevant glenoid bone deficiency, no clinical evidence of pathological anterior inferior laxity (measured with external rotation with the arm at the side inferior to 90 degrees and Gagey sign negative) and arthroscopic finding of isolated anterior Bankart lesion. A total of 20 patients (group A) were randomized to treat Bankart lesion using three bioadsorbable anchors loaded with a #2 braided polyester suture. In 20 randomized patients (group B) two posterior-inferior capsular plications performed with a #1 polidioxanone suture without any capsular shift were added to the same anterior capsulorraphy performed in group A. Postoperative rehabilitation protocol was the same for all 40 patients. Patients were examined preoperatively and at a 2-year follow-up by a single independent expert physician unaware of the surgical procedure. GH ROM, Constant, UCLA and ASES rating scores as well as recurrence of instability were recorded. At follow-up, forward flexion (FF) decreased by a mean value of 14.5 degrees (median -10 degrees ; range -5 degrees to -35 degrees ; P < 0.001) in group B and increased by a mean value of 3.5 degrees (median 0 degrees ; range -25 degrees to 40 degrees ; P < 0.312) in group A; external rotation with arm adducted (ER1) increased by a mean value of 1.8 degrees (median 0 degrees ; range -15 degrees to 30 degrees ; P < 0.924) in group B, and increased by a mean value of 2.6 degrees (median 2.5 degrees ; range -38 degrees to 40 degrees ; P < 0.610) in group A; external rotation with arm abducted at 90 degrees (ER2) decreased by a mean value of 2.9 degrees (median 0 degrees ; range: -20 degrees to 10 degrees ; P < 0.161) in group B and increased by a mean value of 0.7 degrees (median 0 degrees ; range -30 degrees to 25 degrees ; P < 0.837) in group A; the IR2 decreased by a mean value of 2.4 degrees (median -3.5 degrees ; range -15 degrees to 10 degrees ; P < 0.167) in group B and increased by a mean value of 2.2 degrees (median 0 degrees ; range -20 degrees to 30 degrees ; P < 0.456) in group A. The UCLA mean score gains by 43.1% (median 40; P < 0.001) relatively, and of 45.2% relatively (median 40; P < 0.001), respectively, in group B and A, ASES mean score relatively gains by 21.7% (median 21.2%; P < 0.001) in group B, and of 19.2% (median 18.9%; P < 0.001) in group A, and Constant mean score improves by 20.2% (median 16.5; P < 0.001) in group B, and 10.2% (median 8.4%; P < 0.001) in group A. Thus, the only statistical significant differences were the reduction of forward flexion in group B and the improvements of the scores in both groups. No recurrence of instability was found in the plicated group, while in the non-plicated group we had one traumatic recurrence. In conclusion, arthroscopic posterior-inferior plications associated with a Bankart lesion repair in a selected group of patients seem to reduce only FF, without any effect on rotation. A longer follow-up and a larger number of patients are needed to give definitive conclusions on the benefit to the recurrence rate.