【Objective】The technology of using chemicals to control soil erosion is mainly implemented through spraying high molecular compounds over on the surface soil layer to improve its soil structure, to enhance stability of its soil aggregates, to prevent dispersion of its clays and hence to increase rain water infiltration rate and eventually to control surface runoff and soil erosion. Jag S is a kind of neutral polysaccharide and a new type of high polymer. This study on effect of the use of this high polymer on rainwater infiltration rate in loess slopes is intended to lay down a new theoretic basis for applying the technology of using chemicals to control soil erosion to the Loess Plateau in arid and semi-arid regions for erosion control and water conservation. Control of soil erosion with chemicals is a new method, which is developed on the basis of the modern chemical industry and different from the traditional soil and water conservation methods. 【Method】An indoor experiment was conducted in the Rainfall Simulation Hall at the State Key Laboratory of Soil Erosion and Dryland Farming on the Loess Plateau at the Institute of Soil and Water Conservation. A rainfall simulator system with side-sprinklers was used to simulate rainfall events different in intensity ranging from 0.5 to 3.5 mm min^-1. The sprinklers were 16 m high above the experimental field, providing simulated rainfall events over 80%in uniformity and raindrops 0.25 to 0.375mm in diameter, when a rainfall event was set at 1 ~ 2 mm min^-1in rainfall intensity. Kinetic energy of the raindrops striking at surface soil of the field ranged from 365 to 847 J h^-1 m^-2. The indoor simulated rainfall experiment was designed to explore relationships of rain water infiltration rate and initiation time of runoff with dosage (1, 3 and 5 g m^-2) of Jag S used relative to slope gradient (10°, 15° and 20°) and rainfall intensity (1.0, 1.5 and 2.0 mm min^-1). 【Result】Results show that compared with the CK slope, all the slopes sprayed with Jag S regardless of rate were lowered in infiltration rate during the initial period of all rainfall events. However, the spraying of 1 and 3 g m^-2 Jag S significantly raised rain water infiltration rate on all loess soil slopes in the late stages of the rainfall events, and reduced the buffering effect during the initial periods of the rainfalls. Consequently, the rain water infiltration rate increased by 21.53% and the infiltration enhancement rate by 9.17%. On the contrary, the spraying of 5 g m^-2 Jag S weakened the rainfall water infiltration rate in all rainfall events except for the one lowest in intensity (1.0 mm min^-1) and in all the loess slopes. The soil slopes sprayed with Jag S, regardless of rate, differed very slightly in runoff initiation time, however, they had runoff initiated earlier than the CK slope. The higher the Jag S spraying rat, the earlier the runoff initiated. Obviously the effects of a high Jag S spraying rate and a high slope gradient on runoff initiation were significant. The runoff initiation advancement rate on slopes sprayed with 1, 3 and 5 g m^-2 was 47.26%, 50.47% and 66.28%, respectively, under rainfalls the same in intensity. 【Conclusion】All the finding indicate that Jag S is to a certain extent, effective to improve soil structure of the loess slopes and increase rainwater infiltration rate and hence reduce soil erosion. Consequently, spraying of the neutral polymer Jag S on surface soil of a loess slope is a potential alternative to control soil erosion on the Loess Plateau. The findings of this study may serve as a scientific basis for soil and water conservation using polymers.