Abstract: Forest and grass vegetation restoration can be one of the most important measures to effectively control soil erosion at present. Four factors have posed a great impact on soil erosion and sediment yield, including the vegetation types, coverage, structure, and distribution patterns. However, the current characterization of vegetation patterns is only limited to the qualitative description. It is still lacking in the quantitative parameters reflecting the vegetation pattern and the impact on erosion and sediment yield. In this study, the quantitative characterization index of vegetation pattern was investigated to establish the quantitative relationship with the sediment yield, particularly for the greening with water, as well as the soil and water conservation function of vegetation. Meanwhile, the connectivity indexes were also proposed to consider the process of water and sediment transport. A new perspective was provided to explore the mechanism of runoff, sediment yield, and transport processes, particularly for the responses to the changes in land use landscape patterns. Taking 40% of Festuca elata grass-covered slope as the research object, an experiment of artificial simulated rainfall was conducted indoors. An investigation was made to explore the changes in the sediment yield process under different slope gradients (5°, 15°, and 25°) and vegetation patterns (bare land, cross-slope band, down-slope band, block-shaped mosaic, and uniform point-shaped pattern). Two classic connectivity indicators were preferred to characterize the hydrological connectivity of the slope, including the Mean Flow Length Index (MFLI) and the Index of Connectivity (IC). Therefore, a systematic analysis was also made to determine the changes in the MFLI and IC under different grass patterns and the quantitative relationships with the slope gradient and erosion amount. The hydrological connectivity indicators were then verified to characterize the effect of vegetation patterns on erosion and sediment yield. The results showed that the regulation effect of grass pattern on the runoff was closely related to the slope gradient. The best runoff reduction was achieved in the block-shaped mosaic grass pattern when the slope gradient was less than 15°. Once the slope gradient increased to 25°, the pattern with the optimal effect was transformed into the cross-slope band pattern. The average erosion amount values of the grass-covered slopes with the down-slope band pattern, uniform point-shaped pattern, block-shaped mosaic pattern, and cross-slope band pattern were 42.9%, 55.7%, 62.4%, and 78.0% lower than that of the bare land pattern, respectively. The cross-slope band pattern performed the optimal effect of erosion resistance and sediment reduction, while the least was the down-slope band pattern. The relationships between the MFLI and IC of different grass pattern slopes were also consistent with the difference in average erosion amount, and the order from minimum to maximum was as follows: cross-slope band pattern, block-shaped mosaic pattern, uniform point-shaped pattern, down-slope band pattern, bare land pattern. The MFLI and IC can be combined with the slope gradient to simulate and evaluate the erosion and sediment yield of slopes with different grass patterns, where the coefficient of determination (R2) can reach 0.84. The IC presented better synergy and less dependence on the slope gradient, indicating an effective and preferred parameter to characterize the effect of vegetation pattern on erosion and sediment yield, compared with the MFLI. The findings can also provide a strong reference to rapidly optimize the forest and vegetation pattern of soil and water conservation in ecologically fragile areas.