Abstract: In the context of large-scale metagenomic data analysis, there is an urgent need for a robust and high-throughput method to rapidly assign phages to their cyanobacterial hosts. Targeting CRISPR spacer sequences in cyanobacterial genomes offers a promising strategy to predict phage–host linkages. In this study, we performed metagenomic sequencing on cyanobacterial bloom samples from experimental ponds and identified genome sequences of Microcystis cyanophages through CRISPR spacer matching, and the results demonstrated the efficiency and reliability of this approach. Across metagenomic datasets of different seasons, the dynamics of Microcystis cyanophages and Microcystis cells exhibited congruent temporal patterns, lending further support to the predicted phage-host associations. Furthermore, comparative genomic analysis of six relatively complete cyanophage genomes revealed that, except for one genome highly similar to the known cyanophage MaMV-CH01, the remaining five represent genomes of novel Microcystis phages.