This work aimed to study the performance of Santa Inês meat lambs receiving different levels of inclusion of cactus pear (Nopalea cochenillifera) in substitution of Tifton (Cynodon dactylon) hay. The trial was conducted using 48 Santa Inês male lambs at 100 d of age, with average initial weight of 18.75 ± 0.46 kg. Diets were formulated to be isoproteic and isoenergetic. Animals had ad libitum access to feed. The experiment was carried out as a completely randomized factorial design (3 × 2), with 3 levels of substitution of Tifton hay with cactus pear (30%, 50% and 70%), subdivided in 2 groups, with or without access of water, and a control treatment. DMI significantly differed among treatments (P < 0.001) and the interaction among levels of cactus and access of water were not significant for all variables (P > 0.05). The equations to estimate DMI and the ADG in the diets with cactus pear were: DMI = 1572.3 – 17.8x – 112.8y + 0.07x² + 1.78xy (R² = 0.63); ADG = 145.3 + 1.2x + 15.5y – 0.02x² – 0.10xy (R² = 0.29), being x = level of inclusion of cactus pear in the diet (30, 50, 70%) and y = animals with (1) or without (0) access of water. In contrast, the equations to predict the DMI and the ADG including the control treatment with the group that had access of water were: DMI = 897.3 + 7.8x - 0.16x² (R² = 0.40); ADG = 109.2 + 3.2x – 0.04x² (R² = 0.43), being x = diet control and levels of cactus (0, 30, 50, 70%). The maximum value of DMI was 1.1 kg/day and the minimum was 0.7 kg/day, obtained for the group receiving 30% of cactus without access of water and 70% without access of water, respectively. The ADG was different among treatments (P < 0.05) ranging between 177.8 g to 106.8 g for the group with 30% of cactus in the diet with access of water and the control treatment, respectively. Feed efficiency was affected only by treatments, being the group that received 70% of cactus in the diet with access of water the most efficient (0.19 g). In conclusion, cactus pear is recommended at 30% (with access of water) or 50% (without access of water) of substitution of Tifton hay, resulting in ADG of 178 or 157 g/day, respectively.