Abstract: The determination of erosive rainfall standard is crucial for soil erosion prediction, which can effectively reduce the workload of prediction and improve the accuracy of prediction as well. Based on the measured rainfall, runoff, and sediment yield data from 2013 to 2024 at the Guangrong Small Watershed Key Water Erosion Observation Station of the National Soil and Water Conservation Monitoring in Hailun City, Heilongjiang Province, the erosive rainfall standard under four land use (bare land, conventional tillage, no-tillage with straw mulching, and wasteland), as well as the number of years required to stabilize the standard was compared with the rainfall erosivity deviation method, and relevant indicators were used to evaluate the accuracy of erosive rainfall standard. The results show that, for bare land plot, the erosive rainfall (P) standard is 12.6 - 13.3 mm, and the erosive maximum 30-minute rainfall intensity (I₃₀) standard is 9.9 - 12.0 mm·h⁻¹; for the conventional tillage plot, the P standard is 14.4 - 18.0 mm, and the I₃₀ standard is 12.7 - 16.5 mm·h⁻¹. The erosive rainfall standards of no-tillage with straw mulching and wasteland plots are both higher than those of the bare land and traditional tillage plots. The P and I₃₀ standards of no-tillage with straw mulching plots are averagely increased by 1.1 and 1.3 times compared with the bare land plot, and by 65.2% and 68.7% compared with the conventional tillage plot, respectively. The P and I₃₀ standards of the wasteland plots are averagely increased by 1.2 and 1.3 times compared with the bare land plot, and by 70.0% and 72.8% compared with the conventional tillage plot, respectively. By comparing four evaluation indicators, namely relative error, wrong selection index, efficiency index, and erosion loss, the I₃₀ standard is suitable for bare land and conventional tillage plots, while the P standard is suitable for no-tillage with straw mulching and wasteland plots. With the increase in rainfall data duration, the relative error and coefficient of variation of the P and I₃₀ standards decrease. When the rainfall data reaches 11 years, the qualification rates of the P and I₃₀ standards for all land use reach 100%, meaning they become stable. This study provides a reference for soil erosion prediction and soil and water loss prevention and control planning in black soil with rolling hills in Northeast China.