The decomposition of plant residue is largely mediated by microorganisms, such as soil fungi and bacteria, which are very sensitive to changes in their soil environments. Numerous studies at the microbial community level have emphasized the influence of residue return on functional capacity and the characterization of phenotypic traits. Furthermore, previous studies have also shown that fungi dominate the early stages of decomposition in semi-natural unfertilized soils. The long-term impact of straw return on soil ecosystems, especially fungal gene abundance and community composition, is unclear. In the present study, based on the analysis of soil properties, we investigated the effect of rice straw return on the population size and community structure of soil fungi under different straw return days, using real-time PCR and PCR-DGGE. Subsequently, the correlation between fungal community and soil environmental factors was analyzed using redundancy analysis (RDA). Treatments in the present study were based on different return days (90, 180, 270, and 360 d), and a soil with no straw return and similar topography was used as the control. Each treatment had three replicates. The copy numbers of soil fungal ITS ranged from 2.61×10⁷ to 6.46×10⁷ per g of dry soil and was greatly influenced by straw return days. The 360-d treatment yielded the highest copy numbers, whereas the 90-d treatment yielded the lowest copy numbers. The diversity indices (H, R, and E) increased significantly with increasing straw decomposition time and reached maximum values under the 360-d treatment. The DGGE patterns suggested that straw return altered the fungal community, and significant differences were observed among the treatments with different return days. Eleven DGGE bands were re-amplified, sequenced, and aligned using BLAST; and phylogenetic analysis revealed that the soil fungal community of the straw return included Zygomycete sp., Pythium salinum, uncultured Sarcosomataceae, Ascobolus stercorarius, Lagenidium giganteum, Penicillium sp., Aspergillus sp., Thermomyces lanuginosus, Aspergillus glaucus, Polymyxa graminis, and Acremonium sp., among which Penicillium sp., Aspergillus sp., and Acremonium sp. are able to degrade cellulose. RDA analysis suggested the 90-d and 180-d treatments were similar and clearly distinct from the 270-d and 360-d treatments along both the first and second ordination axes, which indicated a pronounced difference in the community composition of soil fungi at the late straw return stage and a slight difference in the early stage, respectively. The eigenvalues of the first two axes of the fungal RDA results were 0.309 and 0.144 and accounted for 80.7% of the total eigenvalue, which suggested qualified ordination results. The first two axes of the species-environment relationship from the RDA results explained 74.8% of the total variance. Among the six soil parameters measured, the influence of available K, alkaline-hydrolyzed N, total N were not significant (P<0.05), but soil organic carbon, pH, and available P were the main factors that influenced the variation of fungal community structure and diversity. The results suggested that long-term straw return had a significant impact on soil fungal composition and that soil organic carbon, pH, and available P are important factors in the dynamics of soil fungal communities.