Abstract Consumer demand for vegetarian and vegan dog foods is increasing. These specialty diets may rely on alternative plant-based protein sources such as pulses and pulse protein concentrates instead of animal protein to meet amino acid (AA) requirements. Pulses such as field pea and lentil contain more protein than cereal grains, but also contain anti-nutritional factors (ANF) such as phytate and insoluble dietary fiber (IDF) that have the potential to reduce nutrient digestibility. The objectives were 1) to determine if the addition of enzymes before or after extrusion would increase apparent total tract digestibility (ATTD) of pulse-based extruded vegan dog foods; and 2) to compare ATTD of nutrients and gross energy (GE) between dogs and growing pigs fed the same diets. Two formulations were created: low protein (Low) containing 24% crude protein (CP) including 35% field pea flour and 15% lentil flour, and high protein (High) containing 43% CP including 50% pea protein concentrate and 20% lentil protein concentrate. Each formulation was used to prepare 3 diets: control without enzymes (Con), with enzymes added before extrusion (P; X-115 single screw; Wenger, Sabetha, KS) or enzymes added after extrusion (A) for a total of 6 diets. The enzyme blend contained cellulase (480,000 or 510,000 U/kg of diet), xylanase (360,000 or 680,000 U/kg), protease (960,000 or 1,360,000 U/kg), and phytase (12,000 or 17,000 U/kg) for Low and High diets, respectively. Following extrusion, kibbles were ground and 0.5% TiO2 was added to diets. Ten mix breed adult dogs (1.5 - 2yrs; 21-29 kg BW) were fed in a 6 × 6 replicated incomplete Latin square with 10-d periods. Feces were collected during the last 4-d of each period. The ATTD of organic matter (OM) and CP ranged from 77.7 to 86.3% and 74.6 to 87.8%, respectively. The ATTD of OM, CP, crude fat, GE, and ash did not differ between A and Con diets. Feces concentrations of glucose, xylose, acetic and propionic acid was greater (P < 0.05) for A than Con diets, indicating that enzymes hydrolyzed some IDF that was subsequently fermented by gut microbes. Addition of enzymes after extrusion increased (P < 0.05) feces moisture by 1.5% causing loose stools. Addition of enzymes before extrusion reduced (P < 0.05) ATTD of CP due to overheating during drying after wet enzyme processing. The ATTD of nutrients was greater (P < 0.05) for pigs than dogs; Nevertheless, could be a good predictor (R2 = 0.74 - 0.87) for ATTD of OM, CP, and GE. In conclusion, enzyme addition after extrusion did not increase ATTD for the dogs but did impact feces metabolites and fecal quality. Finally, pigs are a good translational model for testing ATTD of dog foods.