Virus disease is an important pathogenic factor of apple. Viral species identification plays the most important role in the prevention and control of apple virus diseases. Fruit trees are generally propagated through a-long-period grafting, leading to the plants are easily infected by a variety of viruses. Moreover, the viruses infecting fruit trees always remain a low concentration and a big sequence variation, thus it is particularly difficult to be identified. Additionally, the known molecular protocols majorly include PCR and molecular hybridization, both of which are available to identify known viruses based on their nucleic acid sequences while not for those unreported viruses with their sequences still unknown. In order to break this bottleneck, this study provided a dataset obtained by deep sequencing of small RNA (sRNA) from a naturally cultivated apple trees, combined with bioinformatics methods to filter original data, assembled sRNAs into contigs, and subjected to Blast analysis of the resulting contigs with the sequences in NCBI to analyze and determine virus species, which should bring a new breakthrough for the identification of apple viruses, and provide a technological support to determinate virus-free apple seedlings.