Excavating the molecular details of many diverse enzymes from metagenomes remains challenging in agriculture, food, health, and environmental fields. We present a versatile method that accelerates metabolic enzyme discovery for highly selective gene capture in metagenomes using next-generation sequencing. Culture-independent enzyme mining of environmental DNA is based on a set of short identifying degenerate sequences specific for a wide range of enzyme superfamilies, followed by multiplexed DNA barcode sequencing. A strategy of 'focused identification of next-generation sequencing-based definitive enzyme research' enabled us to generate targeted enzyme datasets from metagenomes, resulting in minimal hands-on obtention of high-throughput biological diversity and potential function profiles, without being time-consuming. This method also provided a targeted inventory of predicted proteins and molecular features of metabolic activities from several metagenomic samples. We suggest that the efficiency and sensitivity of this method will accelerate the decryption of microbial diversity and the signature of proteins and their metabolism from environmental samples.