Depleted soil quality, decreased water availability, and increased weed competition constrain spring wheat production in the northern Great Plains. New management systems are necessary for improved crop productivity. The objective of our study was to compare productivity and soil water use of spring wheat (Triticum aestivum L.) in four crop rotations (continuous wheat, wheat–pea [Pisum sativum L.], wheat–forage barley [Hordeum vulgaris L.]–pea, and wheat–forage barley–corn [Zea mays L.]–pea) in two tillage (tilled and no-till) and management systems (conventional and ecological). Conventional management included recommended seed rates, early planting date, and broadcast N fertilization. Ecological management included variable seed rates, delayed planting, banded N fertilization, and increased stubble height. Spring wheat in diversified rotations averaged 35 mm greater preplant soil water content, 37 mm greater water use, 0.8 kg ha–1 mm–1 greater water use efficiency, and 473 kg ha–1 and 817 kg ha–1 greater grain and biomass yields than continuous wheat. Wheat in conventional management averaged 28 fewer heads m–2, 4 additional seed head–1, and 2 mg seed–1 heavier seed weight than wheat under ecological management, resulting in 644 kg ha–1 greater yield. Wheat under ecological management used 8 mm more water, but water use efficiency was 2.6 kg ha–1 mm–1 greater under conventional management. Postharvest soil water content was similar among rotations, tillage, and management systems, suggesting that wheat uses most available soil water. Spring wheat in diversified rotations planted early in the season is more resilient and should confer greater production stability than continuous wheat systems planted late.