Results Average time of surgery was significantly shorter in ARTHRO than OPEN: respectively 110 and 120 minutes. The number of intra-operative complications was six (12.5%) in OPEN and five (8.1%) in ARTHRO. The results were comparable in both groups, with no significant difference between OPEN and ARTHRO group: satisfaction rate -96.8% and 91.9%, shoulder function -92.2% and 90%, Walch-Duplay score -83.9 and 76.7 respecively. A significant difference was reported in Rowe score: 87.8 in OPEN and 78.9 in ARTHRO. Another significant difference was found in the presence of Bsubjective apprehension^-a term referring to the subjective perception of instability with no signs of instability at clinical examination -28.7% in OPEN and 50% in ARTHRO. Range of motion in both groups were comparable, however patients in OPEN had significantly lower loss of external rotation in adduction to the side comparing to the contralateral shoulder: 7°versus 14°in ARTHRO. Recurrence was reported in three cases in each group: 6.2% in OPEN and 4.8% in ARTHRO. A revision surgery was performed in four patients (9.3%) in OPEN and six (9.7%) in ARTHRO. Radiographic evaluation showed a significantly lower rate (5%) of graft healing problems (fracture, non-union and osteolysis) after arthroscopic stabilisation, however a partial osteolysis of the proximal part of the bone block was significantly more frequent (53.5%). Conclusions The arthroscopic Latarjet stabilisation showed satisfactory and comparable results to open procedure. We recommend further investigation and development of arthroscopic technique.Level of evidence: III