CONTEXT: Soybean-maize strip intercropping has been widely promoted in China because of its advantages, including improved land-use efficiency and increased total yield. However, the long-term reliance on yield as the primary indicator to measure the merits of cropping systems cannot reflect the sustainability of agricultural systems. OBJECTIVE: To quantify the interspecific interaction effects under different spatial configurations; analyze the coupling coordination between water and crops; evaluate and select optimal spatial configurations; investigate the water-related mechanisms of intercropping, quantify the net biodiversity effect (NE), complementarity effect, and selection effect across different intercropping patterns, and explored the water-crop coupling coordination using the Coupling Coordination Analysis (CCA) model. METHODS: A two-year field experiment (2023-2024) was conducted to evaluate soybean-maize strip inter-cropping systems using a single-factor randomized complete block design with three replicates. Five treatments were established: three strip intercropping patterns, including four rows of soybean intercropped with two rows of maize (4:2), four rows of soybean intercropped with three rows of maize (4:3), and six rows of soybean intercropped with three rows of maize (6:3) and sole maize, sole soybean. Crop growth, soil water dynamics, yield, NE, and its components were calculated, and the coupling coordination degree of the water-crop system was evaluated using the CCA model. RESULTS: The results showed that at the grain-filling (R3) and maturity stages (R6) of maize, which corresponding to the same sampling times for soybean, soil water storage in the soybean strips was significantly higher than that in the maize strips all three intercropping patterns. NE values for all three intercropping patterns were all greater than 0 in both years. Regarding the selection effect, the 4:3 pattern performed significantly better than other patterns. The selection effect of the 6:3 treatment was be negative in 2024. Over the two years, the coordination degree (T) value of the 4:2 pattern was significantly 74.3% and 81.4% higher that of the 4:3 pattern, respectively. In 2023, the coupling coordination degree (D) value of the 4:2 pattern was significantly 80.2% higher than that of the 4:3 pattern. CONCLUSIONS AND SIGNIFICANCE: The 4:2 pattern was optimal, as it maximized net biodiversity effects dominated by complementarity rather than selection effects. It exhibited the highest water-crop coupling coordination, achieving both high productivity and resource sustainability. This can provide a reference for selecting spatial configurations for soybean-maize strip intercropping.