[Objective] to evaluate the relationship between the widening of clavicle and coracoid process tunnel and the loss of reduction after anatomical reconstruction of coracoclavicular ligament (ACCR) with endobuton graft in patients with coracoclavicular ligament injury. [Methods] from November 2012 to December 2021, we retrospectively analyzed 30 patients with type III-V acromioclavicular joint injury treated with anatomical reconstruction of coracoclavicular ligament. The average age of 30 patients was 42.63 ± 15.8 (range 18.0-59.0 years). The right shoulder was involved in 46.7%. The average clinical follow-up was 12.0 ± 3.8 months (range 8.0-18.0 months). A 4mm drill bit was used to create clavicle and coracoid process tunnels during operation. After operation, PACS software was used to measure the distance of coracoclavicular joint, the width of inner and outer clavicular tunnel and the width of inner and outer coracoid tunnel before and after operation and the last follow-up after operation. VAS score, constant Murley score, Ulca score and recovery exercise were used to evaluate the effect of operation. [results] all 30 patients were followed up. X-ray measurement showed that there was significant difference between the two time points from the distance of coracoclavicular joint (30.97 ± 3.27mm) to the last follow-up (32.43 ± 3.48mm) (P < 0.001), and the width of clavicular tunnel from 4mm to the last follow-up (5.69 ± 1.10mm) (P < 0.001), The width of the coracoid tunnel was 4mm to the last follow-up (5.45 ± 0.95mm), the difference between the two time points was statistically significant (P < 0.001), and the ases score increased from (54.20 ± 11.00) before operation to (97.00 ± 5.30) at the last follow-up (P < 0.01). The constant Murley score increased from (58.40 ± 3.70) before operation to (98.10 ± 1.50) at the last follow-up. The difference between the two time points was statistically significant (P < 0.01); The score of ULCA increased from (15.10 ± 1.80) before operation to (34.20 ± 2.90) at the last follow-up (P < 0.01). All patients had no infection, postoperative stiffness, internal fixation failure, clavicular and coracoid process fractures. [Conclusion] patients with type III-V acromioclavicular joint injury treated with anatomical coracoclavicular ligament reconstruction with endobuton graft showed that the clavicular and coracoid process tunnels were significantly widened during postoperative follow-up, there was no correlation between tunnel widening and reduction loss, and the change of tunnel width had no effect on clinical results.