Shape Memory Alloys (SMAs) are functional materials with rapidly expanding applications in medical devices, aerospace, and smart actuators. Among them, Cu–Zn–Al-based SMAs are cost-effective but their performance is often limited by the formation of non-reversible martensite. This study investigates the influence of Si addition and quenching methods on the microstructure, martensitic transformation, and deformation behavior of Cu–Zn–Al SMAs. Alloys with compositions of Cu–27Zn–2.5Al and Cu–27Zn–2.5Al–0.3Si wt.% were fabricated via gravity casting and homogenized at 850°C for 2 hours. The samples were then betatized at 900°C for 30 minutes and cooled using two quenching methods: Direct Quenching (DQ) and Up Quenching (UQ). The UQ process involved reheating after initial quenching to promote atomic ordering and defect relaxation, followed by cooling in an ethanol+dry ice mixture maintained at –5°C. The results reveal that the Cu–27Zn–2.5Al wt.% alloy undergoes a β → β′ martensitic transformation in both DQ and UQ conditions, with UQ producing a more homogeneous and responsive martensitic structure. In contrast, the addition of 0.3 wt.% Si refines the α-phase grains and stabilizes the α + β phase region, thereby suppressing martensite formation. The Si-containing alloy deforms mainly through plastic slip in the α phase, whereas the Si-free alloy exhibits the typical twinning/detwinning mechanism of SMAs. These findings confirm that the combination of alloy composition and quenching route governs phase transformation and deformation mechanisms in Cu-based SMAs, offering insight for designing low-cost functional materials with tunable shape memory behavior.