Designing appropriate nitrogen (N) application rates and planting densities is crucial for ensuring high-yield and high-quality rice production. The effects of N fertilizer and planting density on the soil microbial community composition and abundance and on root N uptake mechanisms were investigated. In this study, high-nitrogen low-density (HNLD), medium-nitrogen medium-density (MNMD) and low-nitrogen high-density (LNHD) experiments were conducted via the bucket planting method, and plant physiology, dry matter mass and N uptake, root enzyme activities and rhizosphere soil microbial communities were examined via metagenomic sequencing and isotope tracing. The HNLD treatment exhibited the highest N absorption and utilization capacity. Compared to those in the other treatments, the δ15N values and abundance increased by 52.96–63.45% during the maturation stage and by 34.48–48.36% during the heading stage. Compared with those of the other treatments, the Ndff (15N derived from fertilizer) increased by 59.33–60.92% during the maturation stage and by 40.24–47.56% during the heading stage. Actinobacteria, Acidobacteria, Chloroflexi, and Proteobacteria were confirmed to be the dominant bacterial groups among the treatments. Based on these findings, a correlation-oriented model and dual-factor correlation network were constructed, focusing on the variations in root nitrite reductase (NIR) activity, rhizosphere soil sucrase (S-SC) and soil catalase (S-CAT) activities, and Actinobacteria and Proteobacteria communities. Overall, this study provides theoretical support for determining N application rates and planting densities in practical rice production, with the ultimate goal of enhancing root N uptake efficiency.