Nitrilase can catalyze nitrile compounds to generate corresponding carboxylic acids. Nitrilases as promiscuous enzymes can catalyze a variety of nitrile substrates, such as aliphatic nitriles, aromatic nitriles, etc. However, researchers tend to prefer enzymes with high substrate specificity and high catalytic efficiency. In this study, we developed an active pocket remodeling (ALF-scanning) based on modulating the geometry of the nitrilase active pocket to alter substrate preference and improve catalytic efficiency. Using this strategy, combined with site-directed saturation mutagenesis, we successfully obtained 4 mutants with strong aromatic nitrile preference and high catalytic activity, W170G, V198L, M197F, and F202M, respectively. To explore the synergistic relationship of these 4 mutations, we constructed 6 double-combination mutants and 4 triple-combination mutants. By combining mutations, we obtained the synergistically enhanced mutant V198L/W170G, which has a significant preference for aromatic nitrile substrates. Compared with the wild type, its specific activities for 4 aromatic nitrile substrates are increased to 11.10-, 12.10-, 26.25-, and 2.55-fold, respectively. By mechanistic dissection, we found that V198L/W170G introduced a stronger substrate-residue π-alkyl interaction in the active pocket and obtained a larger substrate cavity (225.66 Å3 to 307.58 Å3), making aromatic nitrile substrates more accessible to be catalyzed by the active center. Finally, we conducted experiments to rationally design the substrate preference of 3 other nitrilases based on the substrate preference mechanism and also obtained the corresponding aromatic nitrile substrate preference mutants of these three nitrilases and these mutants with greatly improved catalytic efficiency. Notably, the substrate range of SmNit is widened. IMPORTANCE In this study, the active pocket was largely remodeled based on the ALF-scanning strategy we developed. It is believed that ALF-scanning not only could be employed for substrate preference modification but might also play a role in protein engineering of other enzymatic properties, such as substrate region selectivity and substrate spectrum. In addition, the mechanism of aromatic nitrile substrate adaptation we found is widely applicable to other nitrilases in nature. To a large extent, it could provide a theoretical basis for the rational design of other industrial enzymes.