SUMMARYIn order to enhance the durability of high-performance buckling-restrained braces (BRBs) used in bridge engineering, which are expected to withstand severe earthquakes three times without being replaced, aluminum alloys were employed to manufacture BRBs. A series of low-cycle fatigue tests, including 18 specimens, were conducted to address the low-cycle fatigue strength of the aluminum alloy BRB. Test results of all specimens show that stable hysteretic curves were obtained without overall buckling occurrence. Failure mode of the welded aluminum alloy BRB is obviously affected by the ribs' welding under the variable or constant strain amplitude condition. Therefore, another type of aluminum alloy BRB, the bolt-assembled BRB with or without spot-welded stoppers, is proposed and tested. Results showed that the low-cycle fatigue performance of bolt-assembled BRBs with stoppers improved four to five times compared with welded BRBs. However, the stoppers' spot welding has an adverse effect on the failure mode because the crack, which induced the specimen's failure, initiated from the spot weld toes of the stoppers. Both bolt-assembled BRBs with and without stoppers can meet the cumulative inelastic deformation requirement proposed for high-performance BRBs under the constant strain amplitude, not larger than 2%. In addition, under the variable strain amplitude condition, only the bolt-assembled BRB without stoppers has an excellent cumulative inelastic deformation capacity and sustains two cycles of 2.5% strain amplitude. Finally, recommended Manson-Coffin equations and preliminary cumulative damage formulae for welded and bolt-assembled BRBs are given as the references of the strain-based damage evaluation.