This study aimed to explore the impact paths on soil organic carbon and crop yield of completely or partially substituting chemical N fertilizer with organic fertilizers. A four-year field experiment was conducted and included four treatments: (i) N0, no N fertilization application; (ii) NF, only synthetic N fertilizer application; (iii) 1/2OF, organic fertilizer substituted for 100% of the synthetic N fertilizer, with the total N application amount being equivalent to half that of NF; and (iv) 1/3OF + 2/3NF, organic fertilizer substituted for 1/3 of the synthetic N fertilizer with the total N application amount from organic and synthetic fertilizer being equivalent to that of NF. Soil total organic carbon (TOC), labile organic-carbon fractions (microbial biomass carbon (MBC), dissolved organic carbon (DOC), particulate organic carbon (POC), and easily oxidized organic carbon (EOC)), the carbon pool management index (CPMI), soil aggregated distribution, and water-stable aggregate-associated organic carbon were determined. Structural equation modeling (SEM) was used to clarify the impact paths of TOC and garlic yield changes under different N fertilizer treatments. Results showed that compared with N0 and NF, 1/2OF and 1/3OF + 2/3NF significantly increased TOC contents by 14.1–20.6%. Soil MBC, DOC, and EOC under 1/2OF were significantly higher than under N0, whereas the 1/3OF + 2/3NF treatment had significantly greater POC. The CPMI was improved by organic fertilizer treatment, with 1/2OF treatment being significantly higher than N0 and NF. The proportion of soil aggregate mass with particle sizes >2 mm was significantly greater under N0, while 1/3OF + 2/3NF significantly increased the proportion of particle sizes of 0.5–2 mm. Soil water-stable aggregate-associated organic carbon showed a trend of first increasing and then decreasing, with the largest particle sizes being 1–2 mm. Moreover, organic fertilizer significantly increased soil water-stable aggregate organic carbon compared with N0 and NF. Similarly, the garlic yield increased with organic fertilizer treatment, while 1/3OF + 2/3NF significantly increased the yield by 37.2% and 15.3%, respectively, compared with N0 and NF. Furthermore, SEM analysis indicated that fertilizer regimes could directly affect TOC and labile organic carbon components by affecting aggregate-associated organic carbon. In particular, aggregates with particle sizes of 0.5–2 mm played an important role, indirectly affecting garlic yield and CPMI. These results indicate that organic fertilizer application has the potential to improve soil organic-carbon content and garlic yield; moreover, fully applying organic fertilizer can reduce N fertilizer input while still maintaining an increase in soil organic carbon and crop yield in the short term. However, caution is still needed regarding of the type and quantity of organic fertilizer added in different cropping systems, and with different soil textures.