Abstract Purpose To determine whether arthroscopic Latarjet procedure or arthroscopic bony Bankart repair provide better outcomes in terms of rates of recurrent instability, non‐union and complications, as well as clinical scores and range of motion. Methods An electronic literature search was performed using PubMed, Embase®, and Cochrane databases, applying the following keywords: “Arthroscopic bony Bankart” OR “Arthroscopic osseous Bankart” AND “Arthroscopic Latarjet” OR “Arthroscopic coracoid bone block”. Results The systematic search returned 1465 records, of which 29 were included (arthroscopic bony Bankart repair, n = 16; arthroscopic Latarjet, n = 13). 37 datasets were included for data extraction, on 1483 shoulders. Compared to arthroscopic Latarjet, arthroscopic bony Bankart repair had significantly higher instability rates (0.14; CI 0.10–0.18; vs 0.04; CI 0.02–0.06), significantly lower union rates (0.63; CI 0.28–0.91 vs 0.98; CI 0.93–1.00), and significantly lower pain on VAS (0.42; CI 0.17–0.67 vs 1.17; CI 0.96–1.38). There were no significant differences in preoperative glenoid bone loss, follow‐up, complication rate, ROWE score, ASES score, external rotation, and anterior forward elevation between arthroscopic Latarjet and arthroscopic bony Bankart repair. Conclusion Compared to arthroscopic Latarjet, arthroscopic bony Bankart repair results in significantly (i) higher rates of recurrent instability (14% vs 4%), (ii) lower union rates (63% vs 98%), but (iii) slightly lower pain on VAS (0.45 vs 1.17). There were no differences in complication rates, clinical scores, or postoperative ranges of motion. Level of evidence IV.