Abstract: The upland red soil in Jiangxi Province is rich in light, heat, and water resources and provides superior natural conditions for agricultural development. To screen the maize-soybean intercropping system with obvious advantages, better group yield, and economic benefits in the upland red soil of Jiangxi Province, seven treatments were set up in 2018 and 2019, of which five were maize-soybean intercropping systems with 1) 2 m bandwidth and 2-2 rows ratio, 2) 2.4 m bandwidth and 2-3 rows ratio, 2.4 m bandwidth and 2-4 rows ratio, 2.8 m bandwidth and 2-3 rows ratio, and 2.8 m bandwidth and 2-4 rows ratio; other two were sole maize and sole soybean. The effects of different intercropping patterns of maize and soybean were analyzed for crop yield and the interspecific relationships. The results showed that 1) the yield of sole maize and soybean was significantly higher than that of the intercropped ones. In the 2-year experiment, the average yield was 6801.88 kg·hm^-2 under the intercropping treatment of 2.4 m bandwidth and 2-4 rows ratio, which was 7.56% higher than that under the intercropping treatment of 2.4 m bandwidth and 2-3 rows ratio, indicating that increasing one row soybean of the intercropping system with suitable bandwidth benefited yield increase. The average economic benefit of intercropping system with a 2.4 m bandwidth and 2-4 rows ratio was 15 822.96 ￥·hm^-2 for the 2 years, 9.90% and 209.48% higher than that of sole maize and sole soybean, respectively. This indicated that intercropping improved the farmers' economic benefit. 2) In 2019, the land equivalent ratio (LER) of the intercropping treatment with 2.4 m bandwidth and 2-4 rows ratio was 1.77, significantly higher than the other intercropping treatments. The 2-year average LER in this intercropping system was 1.59, compared with 1.55 for the 2.0 m bandwidth and 2-2 rows ratio, 1.44 for the 2.4 m bandwidth and 2-3 rows ratio, 1.35 for the 2.8 m bandwidth and 2-3 rows ratio, and 1.44 for the 2.8 m bandwidth and 2-4 row ratio, which showed better land productivity. 3) When the bandwidth was 2.4 m or 2.8 m, adding one more row of soybean increased the LER by 10.42% and 7.41%, the maize actual yield loss (AYL_M) by 77.01% and 59.02%, the maize invasion (A_M) by 91.89% and 82.22%, the maize competition ratio (CR_M) by 38.69% and 24.11%, the economic benefit (EB) by 8.38% and 4.80%, and the maize harvest index (HI) by 6.12% and 6.25%, respectively. These results showed that increasing the row ratio of soybean could enhance the competitive advantage of intercropped maize with a suitable bandwidth, which improved crop yield and economic benefits. When the bandwidth increased from 2.4 m to 2.8 m with the rows ratios of 2-3 and 2-4, LER decreased by 6.67% and 9.66%, AYL_M decreased by 42.62% and 59.79%, A_M decreased by 64.44% and 73.17%, CR_M decreased by 19.15% and 33.14%, EB decreased by 10.23% and 13.99%, and HI decreased by 2.08% and 0.51%, respectively. These results showed that increased bandwidth weakens the competitive advantage of intercropped maize with different rows ratios and reduced the group yield and economic benefit. In conclusion, the maize-soybean intercropping system of 2.4 m bandwidth with a 2-4 rows ratio better coordinated the crop interspecific relationships and resulted in higher crop yield and better economic benefits. This study provides a reference for the maize-soybean intercropping system in the upland red soil of Jiangxi Province.