Microbial interactions evolve simultaneously with the evolution of microorganisms. Although building Earth co-occurrence network offers valuable insights into the complex connectivity of the global microbiome, the evolutionary feature of these networks remains largely uncharacterized. Here, we simulated the evolution of the Earth microbial co-occurrence network using the Earth Microbiome Project datasets. We found that the Earth microbial co-occurrence network evolved following the Bianconi-Barabasi model, suggesting that inherent traits of microbial taxa shaped potential interaction patterns in microbial communities. The estimated degree growth exponent (β) was the major determinant of accumulated degree of taxa. The positive correlation between β and genome sizes suggested that genome sizes regulated the evolution of the Earth microbial co-occurrence network. The gamma distribution of β suggested that the taxa with low topological fitness were prone to extinction, even though link acquisition by hub nodes was not affected by node extinction and decay. This study provides a framework for predicting evolution trends of microbial co-occurrence network and sheds light on the evolutionary features of the Earth microbial co-occurrence network.