Objective To compare the biomechanical behavior of a triple Endobutton technique for anatomic reconstruction of coracoclavicular ligament and with a modified Weaver-Dunn procedure. Methods Twelve fresh frozen cadaveric shoulders were applied with 70 N in superior, anterior and posterior direction, respectively, to measure displacement of the acromioclavicular joint. The failure test with the load at the rate of 25 mm/min was conducted to record the failure load and failure mode. The specimens were then randomly assigned to 2 groups: the triple Endobutton technique group and the modified Weaver-Dunn procedure group for reconstruction and to conduct displacement test and failure test again. The stability and mechanical strength of acromioclavicular joints after reconstruction under different states were then compared. Results The triple Endobutton technique group had significantly less anterior ((8.72±1.41) mm vs (37.03±5.05) mm) and posterior ((8.03±3.68) mm vs (14.85±1.89) mm) displacement than that in the modified Weaver-Dunn procedure group after reconstruction (P<0.05), and the former had similar displacement ((7.81±2.22) mm anterior and (7.16±1.95) mm posterior) as compared to the intact state. There were no significant differences in superior displacement among the groups. The modified Weaver-Dunn procedure group had significantly smaller failure loads ((172±9) N) than that in the triple Endobutton technique group ((687±115) N) and the intact ligament group ((685±234) N) (P<0.05). Conclusions The triple Endobutton technique has less anterior and posterior displacement and its stability is more closely approximate to the intact ligament; meanwhile, it has similar strength as coracoclavicular ligaments, which can better restore the function of coracoclavicular ligaments.