Improving soil quality while achieving higher productivity is the major challenge in the agricultural industry. Wheat (Triticum aestivum L.)–maize (Zea mays L.) (W–M) rotation is the dominant planting pattern in the Huang-Huai-Hai Plain and is important for food security in China. However, the soil quality is deteriorating due to the W–M rotation’s long-term, intensive, and continuous cultivation. Introducing legumes into the W–M rotation system may be an effective way to improve soil quality. In this study, we aimed to verify this hypothesis by exploring efficient planting systems (wheat–peanut (Arachis hypogaea L.) (W–P) rotation and wheat rotated with maize and peanut intercropping (W–M/P)) to achieve higher agricultural production in the Huang-Huai-Hai Plain. Using traditional W–M rotation as the control, we evaluated crop productivity, net returns, soil microorganisms (SMs), and soil organic carbon (SOC) fractions for three consecutive years. The results indicated that wheat yields were significantly increased under W–P and W–M/P (382.5–579.0 and 179.8–513.1 kg ha−1, respectively) compared with W–M. W–P
and W–M/P provided significantly higher net returns (58.2 and 70.4%, respectively) than W–M. W–M/P and W–M retained the SOC stock more efficiently than W–P, increasing by 25.46–31.03 and 14.47–27.64%, respectively, in the 0–20 cm soil layer. Compared with W–M, W–M/P improved labile carbon fractions; the sensitivity index of potentially mineralizable carbon, microbial biomass carbon (MBC), and dissolved organic carbon was 31.5, 96.5–157.2, and 17.8% in 20–40, 10–40, and 10–20 cm soil layers, respectively. The bacterial community composition and bacteria function were altered as per the soil depth and planting pattern. W–M/P and W–M exhibited similar bacterial community composition and function in 0–20 and 20–40 cm soil layers. Compared with W–P, a higher abundance of functional genes, namely, contains mobile elements and stress-tolerant, and a lower abundance of genes, namely, potentially pathogenic, were observed in the 10–20 cm soil layer of W–M and the 0–20 cm soil layer of W–M/P. SOC and MBC were the main factors affecting soil bacterial communities, positively correlated with Sphingomonadales and Gemmatimonadales and negatively correlated with Blastocatellales. Organic input was the main factor affecting SOC and SMs, which exhibited feedback effects on crop productivity. In summary, W–M/P improved productivity, net returns, and SOC pool compared with traditional W–M rotation systems, and it is recommended that plant–soil–microbial interactions be considered while designing high-yield cropping systems.