Objective To compare the stability of greater tuberosity fractures of humerus treated by three different fixation techniques (screws, tension band, locking plate, respectively) through biomechanical testing, so as to provide the biomechanics basis for choosing a better fixation in the clinical treatment for greater tuberosity fractures of humerus. Methods Standardized fracture models of the greater tuberosity from 18 fresh-frozen proximal humeri with intact rotator cuffs were created. The specimens were randomly assigned to 3 groups and treated by screws, tension band and locking plates, respectively. An increasing force was applied to the supraspinatus tendon. The force displacement curve and two parameters: LtYP(Load to 5 mm yield point) and Ltf(load to failure) were recorded. Results LtYP from the screw group, tension band group and locking plate group was (377±86), (499±90), (793±52) N, respectively, with significant differences among the three groups (P<0.01). Significant differences were also found between the groups as locking plate group (only 3 cases in locking plate group reached to 5 mm displacement before LtF in this study ) and screw group, locking plate group and tension band group, tension band group and screw group (P<0.01). LtF of screw group, tension band group and locking plate group was (744±112), (908±93), (979±143) N, respectively, showing significant differences among them, and which were also found between locking plate group and screw group, tension band group and screw group (P<0.01), but no significant differences were found between locking plate group and tension band group (P>0.05). Conclusions Locking plates show more obvious biomechanical stability than screws and tension band, which provides a new and better choice for treatment of isolated greater tuberosity fractures of humerus.