Purpose A systematic review to determine the effect of different types of joint preparation, joint fixation, and postoperative weight-bearing protocols on non-union frequency in first metatarsophalangeal joint (MTPJ) arthrodesis in patients with moderate-to-severe hallux valgus. Material and methods A systematic literature search (PubMed and EMBASE), adhering to PRISMA guidelines. Data on MTPJ preparation, fixation, weight-bearing, and non-union in patients with moderate-to-severe hallux valgus were collected. Quality assessment was performed using the Coleman Methodology Score. Results Sixteen studies (934 feet) were included, generally of medium quality. Overall non-union rate was 7.7%. At 6.3%, convex/concave joint preparation had the lowest non-union rate vs 12.2% for hand instruments and 22.2% for planar cuts. Non-union of 2.8% was found for joint fixation with a plate combined with a lag screw vs 6.5% for plate fixation, 11.1% for crossed screw fixation, and 12.5% for a plate with a cross plate compression screw. A 5.1% non-union frequency was found following postoperative full weight-bearing on a flat shoe vs 9.3% for full weight-bearing on a heel weight-bearing shoe and 0% for a partial weight-bearing regimen. Conclusion Based on medium-quality papers, joint preparation with convex/concave reamers and joint fixation with a plate using a lag screw show the lowest non-union rate. Full postoperative weight-bearing in a stiff-soled postoperative shoe is safe and not associated with non-union vs a more protective load-bearing regimen. Further research should focus on larger sample sizes, longer follow-ups, and stronger study designs.