This paper probes into the molecular genetic mechanism of the formation of species, subspecies and variety in evolving progression, and brings forward 5 criteria of an ideal strategy in species identification: stating the specific characteristics at species, subspecies and variety level without any interference of too high polymorphism at individual or population level; keys should be distributed as 0 or 1, e. g. yes or no; satisfying repeatability and simple operation; high veracity and reliability; adaptability to widely various specimen. Respectively, this paper reviews two strategies focusing on detecting the fragment length polymorphism and base replacement and lays out some detail methods under above strategies. It demonstrates that it is not possible to solve all species problems by pursuing identification with only a single gene or DNA fragment. Only based on thorough consideration of all strategies, a method or combined several methods could bring satisfying reliability. For advanced focuses, it requires not only development and optimization of methods under above strategies, but also new originality of creative strategies.