【Objective】Gully erosion is very severe in the Yuanmou Dry-hot Valley Region, dominated with headward erosion, which is characterized by steep gully headcuts. During the field investigations, headcut height is often regarded as one of the main indices of distinguishing the type of gully head activity. The abrupt drop of concentrated flow at headcut would cause a sudden change of the runoff in hydrodynamics, thus accelerating the headward erosion processes. It is, therefore, of great significance to carry out the preliminary study on effects of headcut height on characteristics of soil erosion and sediment yield in the gully headward processes, to the gully erosion research in this region as scientific supplement.【Method】To simulate different step heights, based on an in-situ active bank gully head, two experimental plots were constructed with the headcut height of 25 cm and 50 cm, separately. And in November 2015 a series of in-situ scouring experiments were carried out with the flow set at 120 Lmin^-1. The entire scouring test lasted 140 min, consisting of 5 test runs. During the scouring tests, cross-sections were set in the drainage area and along the gully bed at1 m intervals, for observation of changes of runoff in width, depth and velocity at regular intervals. Meanwhile, sediment samples were also collected at regular time intervals at the outlet of the drainage area and the plots, separately, for analysis of sediment content with the oven drying method. Besides, before and after the experiment, topographic data of the two plots were collected with the high-resolution close range digital photogrammetry measurements, for analysis of changes in landform, morphological parameters and net soil loss.【Result】Results showed that: (1)The scouring power of jet flow differed significantly between the two plots. When surface runoff turned into impinging jets, the mean potential energy transformation in the two plots was about 4.89and 9.78Js^-1, separately with shear stress reaching 25.9 and 53.5 Pa, on average respectively, both in multiple relations. And the velocity at the instance when the flow touched the bottom of the headcut reached as high as 2.28 and 3.18 ms^-1, separately, which was 4.1 and 5.6 times larger than that at the brink point respectively. (2) The two plots also differed significantly in morphological characteristics of plunge pools and soil erosion volume at lower end. At the end of the experiment, the incision depth of jet flow reached7.38 and 7.50 cm in maximum, and was 3.87 and 5.16 cm on average, separately, in the two plots. Especially, the growth of incision depth with the experiment going on varied sharply between the two plots. Furthermore, during the scouring tests, the soil erosion cumulative contribution rate at the gully headwall and in the bed at the lower end reached as high as45.1% and 64.9%, and was 41% and 54%on average. (3) The two plots also differed sharply in runoff sediment concentration (p< 0.1). Excluding the maximum concentration at the beginning of each test, the mean sediment concentration was 7.51 and 18.76 g L^-1, separately, about 1.5 times in difference. Especially, during the first previous 80 min of the scouring tests, when runoff flew over the headcut, the runoff sediment concentration increased by 25.28 and 108.78 g L^-1 in maximum and by 6.25 and 25.49 g L^-1on average, separately, in the two plots. 【Conclusion】This study has tentatively concluded that the effect of gully headcut height affecting runoff scouring force and characteristics of soil erosion and sediment yield is probably the key factor influencing the headward erosion process. However, it is still not very clear about how the effect would vary in the case of higher headcuts. More experiments with a variety of headcut heights should be conducted, in an attempt to supplement more complete meaningful theories to the study on dynamics and mechanisms of gully headward erosion in dry-hot valley regions.