Attention has recently been paid to primary arthroscopic repair to treat ACL tears because of the disadvantages associated with reconstruction. However, there remain many unanswered questions and concerns about its application in the treatment of ACL tears. (1) Does primary arthroscopic repair using the pulley technique result in satisfactory ROM (a functional ROM with a flexion contracture of 30° or less), knee stability, and functional scores in patients with partial proximal ACL tears? (2) What complications are associated with primary arthroscopic repair using the pulley technique in patients with partial proximal ACL tears? Between January 2014 and March 2016, we treated 23 patients surgically who had partial proximal ACL tears and excellent tissue quality (defined as a remnant with mild interstitial tearing and the ability to hold sutures). All patients meeting those two criteria were treated using primary arthroscopic repair using the pulley technique. During that period, this represented 13% (23 of 183) of the patients we treated surgically for an ACL tear. Patients were excluded if they had other ACL tear types, insufficient tissue quality (defined as a severely torn remnant that was not strong enough to hold sutures), multi-ligamentous injuries, or substantial arthrosis (chondromalacia greater than Outerbridge grade 3, most of which underwent conversion to ACL reconstruction). Clinical outcomes were assessed using ROM, the anterior drawer test, the Lachman test, Lysholm score, Tegner activity score, IKDC subjective score, and radiographs. Twenty-one patients were observed for a mean (range) period of 36 months (25-49), and two were lost to follow-up. At the most-recent follow-up examination, all patients achieved full extension and only one patient lacked full flexion, with a flexion contracture of 10°. Twenty patients had no instability on the anterior drawer test and Lachman test findings, and one patient had a 1 + anterior drawer test. The mean Lysholm score improved from a mean ± SD of 71 ± 9 before surgery to 94 ± 6 (mean difference 23 points [95% CI 20 to 25]; p < 0.001) at latest follow-up. The IKDC subjective score improved from 64 ± 10 to 86 ± 11 points (mean difference 22 points; p < 0.001). We found no difference in the Tegner score from before surgery to latest follow-up (6.3 ± 1.2 versus 6.1 ± 1.2; mean difference 0.2; p = 0.056). One patient re-ruptured his ACL 2 months after surgery in military training during an obstacle race. No complications such as infection, thrombosis, stiffness, patellofemoral pain, or implant failure were observed. Primary arthroscopic repair using the pulley technique can achieve short-term clinical success in a carefully selected (the selection process includes first identifying the ACL injury pattern preoperatively with MRI, then confirming the diagnosis under arthroscopy, and deciding whether to perform a repair intraoperatively) subset of patients with partial proximal ACL tears and excellent tissue quality (defined as a remnant with mild interstitial tearing and the ability to hold sutures). Despite the promising clinical outcomes of our study, this technique should not be widely adopted unless it has been compared directly with ACL reconstruction, so future studies should be conducted to compare the clinical outcomes between this technique and ACL reconstruction, and longer-term follow-up is necessary to identify whether there is deterioration in the clinical outcomes over time. Level IV, therapeutic study.