Abstract: Ruichang yam (Dioscorea opposita), a characteristic economic crop in Jiangxi Province, faces severe challenges due to continuous cropping obstacles, which hinder sustainable production by causing soil degradation, pest accumulation, and yield reduction. To address these issues, a field experiment was conducted from 2022 to 2024 in a typical red soil dryland area of Jiangxi Province, aiming to identify efficient and sustainable cropping patterns that mitigate continuous cropping obstacles and enhance comprehensive system benefits. Four multiple cropping patterns were designed: Chinese milk vetch-yam (CK, control), ryegrass-yam||spring maize-mung bean (B), hairy vetch-yam||soybean-autumn maize (C), and rape-yam||peanut-sesame (D). The grey relational analysis method was employed to evaluate crop productivity and comprehensive benefits across economic, ecological, and social dimensions. The results demonstrated that multiple cropping patterns (B, C, and D) significantly outperformed the CK pattern in terms of economic yield, straw yield, and total biomass, with increases ranging from 3.14 to 8.16 times, 2.89 to 3.75 times, and 2.97 to 5.14 times, respectively. Over the two-year study, the C pattern (hairy vetch-yam||soybean-autumn maize) exhibited the highest weighted correlation degrees in economic benefits (0.404), ecological benefits (0.312), and social benefits (0.152). Specifically, the C pattern achieved the highest values in key indicators including total output value, net output value, output-input ratio, profit rate of agricultural expenses, straw incorporation amount, soil-improving crop index, crop diversity index, soil organic matter content, labor productivity, and grain and oil yield. These results indicate that the C pattern effectively optimized resource allocation, enhanced system productivity, and improved sustainability. The superior performance of the C pattern can be attributed to its rational species combination and efficient spatiotemporal arrangement. The legume-cereal intercrop system (soybean and autumn maize) facilitated biological nitrogen fixation and improved light and temperature utilization, thereby increasing both economic and ecological returns. Additionally, the incorporation of hairy vetch as a green manure further enhanced soil fertility and structure. This pattern not significantly alleviated continuous cropping obstacles but also ensured high and stable yield, making it a suitable strategy for improving land productivity and promoting green agricultural development in red soil drylands. In conclusion, the hairy vetch-yam||soybean-autumn maize model can achieve “high yield-high efficiency-sustainable” collaborative optimization through efficient allocation of resources, which is a suitable model to solve the continuous cropping obstacle of yam in red soil dryland and improve the comprehensive productivity of land, and can provide technical options for the green development of regional dryland agriculture.