Through simulated rainfall experiments, the experimental micro-plots were used to simulate different soil macropore structures (exposed type, shallow-buried type, and deep-buried type), and the totally soil filled micro-plots were used as a control to discuss the effects of soil macropore structure on karst soil and water loss processes. The results showed that: Compared with the totally soil filled micro-plots, the presence of soil macropore significantly accelerated the vertical water leakage, and different soil macropore structures had distinction runoff peaks (exposed > shallow-buried > deep-buried > totally soil filled). The soil macropores also exacerbated soil loss risk, and there were significant differences in soil loss between different soil macropore structure treatments (exposed > shallow-buried > deep-buried > totally soil filled). The soil loss amounts were relatively small (less than 10 g), and larger soil loss amounts only occurred under extreme conditions (exposed soil micropores with a rain intensity of 120 mm/h). Small-sized sediments less than 2 mm were dominated during soil loss processes for all treatments, and the sediments larger than 5 mm mostly occurred in the exposed soil macropore treatments. There were significant correlations between soil loss factors (soil loss amount and fine sediment content) and water leakage factors (water leakage initiation time and water leakage amount).The water leakage process was mainly affected by rainfall intensity, and the soil loss process was mainly affected by the soil macropore structures. The results of this study can provide scientific and technological support for the prevention and control of rocky desertification and groundwater pollution in karst areas.