The combination of forage-crop rotation with conservation tillage has long been known to result in proven productivity and sustainability, but less information is available on the influence of long-term conservation tillage practices on the movement of soil water and residual nitrate at deep soil depths in forage-crop rotation systems. In this study, we measured the changes in soil water storage, residual soil nitrate accumulation and their relationship to study the long-term effects of tillage and mulching practices on regional nitrogen management in maize (Zea mays L.)-wheat (Triticum aestivum L.)-common vetch (Vicia sativa L., annual legume forage) rotation system. Soils were collected from the 500-cm soil profile at the harvest stage of the maize and common vetch after 19 years of continuous conventional tillage (T), conventional tillage followed by straw mulching (TS), no tillage (NT), and no tillage followed by straw mulching (NTS) on the Loess Plateau, China. The results showed that the NT practices increased soil water storage by 3.07% compared with the T practices in the maize field, mulching practices increased soil water storage by 3.47% compared with no mulching practices in the common vetch field, and the improvement effect was mainly concentrated in the deep soil layer (150–500 cm). When compared with no mulching, straw mulching practices increased residual soil nitrate accumulation by 51.20 kg ha⁻¹ at shallow soil depths (0–120 cm) but decreased this accumulation by 206.09 kg ha⁻¹ at deep soil depths (120–500 cm) in the maize field; however, straw mulching decreased nitrate accumulation by 16.25 kg ha⁻¹ throughout the whole profile (0–500 cm) in the common vetch field. The NT practices decreased residual soil nitrate accumulation by 131.65 kg ha⁻¹ compared with the T practices in the maize field but had no effects on the common vetch field in the 0–500-cm soil profile. Greater changes in soil residual nitrate accumulation appeared at depths of 200–500 cm than at other depths. Moreover, according to structural equation model, nitrite accumulation was limited by soil water storage. Soil nitrate peaked at deeper soil depths in the NT (90–120 cm) and TS (250–300 cm) treatments than in the T treatment (60–90 cm), whereas the NTS treatment resulted in no nitrate accumulation peak across the whole soil profile. Soil water recharge and depletion were deeper in the NT and TS treatments than in the T treatment, and this effect could limit the further leaching of soil nitrate into deep soil. The depth of soil water recharge was deeper than the depth of nitrate accumulation, and the vertical movement speed of soil residual nitrate was slower than that of soil water. Therefore, the downward movement of soil nitrate lagged behind that of soil water. Together, these results demonstrated that the NTS treatment exerted the strongest effect on maintaining soil water and decreasing residual nitrate accumulation at deep soil depths, which will be favorable for improving regional N management and assessing N budgets in agroecological systems on the Loess Plateau.