The aim was to evaluate the chemical composition, carbohydrates, protein fractionation and in vitro gas production of silages composed of spineless cactus and tropical forages and their effect on sheep performance. Treatments consisted of silages: corn silage (CS), spineless cactus silage (SCS), spineless cactus + gliricidia (SCG), spineless cactus + buffel grass silage (SCBG) and spineless cactus + pornunça (SCP). Silos were opened 60 days after ensiling, and analyses were carried out. The digestibility test lasted for 36 days, with eight animals per treatment. A completely randomized design was adopted. Considering carbohydrate fractionation, CS, SCS and SCBG silages had higher total carbohydrate content (p = 0.001). The SCS silage presented a higher A + B1 fraction (p = 0.001). The SCBG and SCG silages showed a higher B2 fraction (p < 0.0001) compared to the CS and SCS silages. The SCBG and SCP silages presented a higher C fraction (p = 0.001). For protein fractionation, the SCP and SCG silages showed higher crude protein contents (p = 0.001). The CS and SCS silages showed a higher A fraction (p = 0.001). The SCBG silage presented a higher B1 + B2 fraction (p = 0.001). The SCG silage showed a higher B3 fraction (p = 0.006) compared to SCBG silage. The SCS and SCP silages showed a higher C fraction (p = 0.001). Exclusive SCS silage showed higher in vitro dry matter digestibility (p = 0.001), dry matter degradability (p = 0.001) and total gas production (p = 0.001). The use of the SCBG, SCP and SCG silages to feed sheep increased the dry matter intake (p < 0.001). Sheep fed the SCG silage showed greater dry matter and crude protein digestibility compared to the sheep fed the CS, SCS and SCP silages (p = 0.002). There was a higher water intake (p < 0.001) with the use of the SCS and SCG silages to feed the sheep. The SCP and SCG silages provided a greater intake (p < 0.001) and excretion (p < 0.001) of nitrogen by the animals. Although there were no differences between the treatments for daily gains, lambs that received the spineless cactus-based silage associated with tropical forages showed higher gains (160-190 g/day) than lambs that received CS silage (130 g/day). Thus, the use of spineless cactus associated with buffelgrass, pornunça and gliricidia to prepare mixed silages (60:40) to feed sheep has potential use to feed sheep, with positive effects on nutrient degradation and increases in dry matter intake. Under experimental conditions, we recommend the exclusive use of spineless cactus silage associated with buffel grass, pornunça and gliricidia in feeding sheep in semi-arid regions, as it provides nutrients, water and greater daily gains compared to corn silage.