Abstract: To address structural application limitations due to performance deficiencies of recycled concrete, multi-scale axial compression tests on steel fiber recycled aggregate concrete (SFRAC) specimens (100~200 mm in size) were conducted. Combined with acoustic emission monitoring, the coupled effects of the steel fiber volume content (0%~2%), of the recycled aggregate replacement ratio (0%~100%), and of the specimen size on the compressive performance were investigated. Through acoustic emission characteristic parameters, the effects of mix proportion parameters and of size effect on the compressive damage evolution process were revealed at the micro-level. A quantitative model between acoustic emission activity N and stress level σ was established, correlating microscopic damage with macroscopic behaviors. Based on a Box-Behnken design, prediction models for the axial compressive strength and the toughness index considering interaction effects were developed. Multi-objective optimization was performed using a response surface methodology, and an economic mix design method was proposed by introducing a fiber reinforcement efficiency function. This study provides experimental evidences and an optimization approach for the design and the engineering application of high-performance green SFRAC.