Abstract:[Objective] Vegetation restoration can effectively prevent and control soil and water losses at abandoned production sites and construction projects. Determining the soil and water conservation effects of various vegetation restoration methods in an abandoned dreg field can provide a theoretical basis for subsequent management and supervision of the area. [Methods] An artificial rainfall simulation experiment was set up to obtain the runoff and sediment production characteristics of a slope surface covered with selected waste slags under different planting methods (broadcast sowing, drill sowing, and hole sowing) and vegetation restoration stages (growth, maturity, and withering periods) from the production and construction projects in the Dabie Mountains of Western Anhui Province, and the patterns and differences were analyzed by methods such as cumulative average deviation. [Results] Compared with bare slope, vegetation restoration effectively delayed the initial runoff generation time. The most obvious effect was observed for broadcast sowing. Various planting methods exhibited varying degrees of effectiveness in reducing runoff and sediment, with the overall runoff and sediment output from a slope following the order of bare slope > hole sowing > drill sowing > broadcast sowing, and withering period < maturity period < growth period. The production of sediment and runoff were characterized by distinct processes, and the effect of reducing runoff was weaker than that of reducing sediment. [Conclusion] The runoff and sediment reduction benefits of planting methods were shown to be greater for broadcast sowing than for drill sowing, with hole sowing being the worst. The efficacy of reducing runoff and sediment during the growth period surpassed that of the maturity period, while the withering period yielded the least benefits. Among the various sowing methods, broadcast sowing exhibited the highest efficiency in reducing runoff and sediment during the growth period, with rates of 49.6% and 95.5% respectively.