Abstract: Abstract: Near-surface vegetation and micro-topography have a great impact on the surface runoff process in the mountainous areas of southwestern China. It is, therefore, necessary to clarify the combined influence on the soil and water loss of steep slopes. In this study, 10 natural runoff plots were set to reveal the effects of cypress spatial pattern and micro-topography on the surface runoff. Pearson correlation coefficient method was applied to determine the correlation between cypress spatial pattern, micro-topography, and characteristic parameters of surface runoff. A Response Surface Method (RSM) was also utilized to analyze the measured cypress spatial pattern and micro-topography in the process of precipitation and surface runoff. The results showed that: 1) The peak flow coefficient in all runoff plots behaved short-duration rainstorm (the rainfall lasted no more than 3 h, and the maximum 1 h rainfall intensity exceeded 30 mm/h) > long-duration rainstorm (the rainfall lasted more than 3 h, and the average rainfall intensity exceeded 2.5 mm/h) > long-duration heavy rainfall (the rainfall lasted more than 3 h, and the average rainfall intensity was between 1.5-2.5 mm/h), indicating that the blocking effect of different underlays on surface runoff decreased, with the increase of rainfall intensity and concentration of precipitation. 2) Five factors were significantly correlated with the peak flow (P<0.05) in the long-duration heavy rainfall or rainstorm, including topographic relief, surface roughness, runoff path density, contagion index of cypress, and stand density of cypress. Nevertheless, there was no significant correlation with the peak flow under the condition of short-duration rainstorm. 3) In long-duration heavy rainfall or rainstorm, the response of peak flow to the composite index were that: a) There was no significant change in the peak flow, when the composite index for cypress spatial pattern (V) was below 20.5, while the composite index of micro-topography (U) was below 10.5 (Low surface roughness, high runoff path density, and low stand density of cypress). b) The peak flow was relatively promoted, as the composite index for cypress spatial pattern increased within a certain range, when U > 9.0 (High surface roughness, low runoff path density, low stand density of cypress, and uniform structure among cypress). c) The peak flow was reduced significantly with the increase of V value, when U < 7.5 and V > 18 (Low runoff path density, while high stand density of cypress). As such, when the V value was adjusted to 41 under the condition of long-duration heavy rainfall (rainstorm), an optimal combination was achieved, where the peak flow was reduced from 39.6 L/10 min (14.2 L/10 min) to 6.3 L/10 min (4.2 L/10 min), while the reduction rate reached 84% (70%). d) Once the V value exceeded the critical one (the specific critical value increased with the increase of U value), the dominant influencing factor of peak flow shifted gradually from cypress spatial pattern to micro-topography. This finding can provide promising theoretical support to accurately adjust the vegetation patterns for the prevention and control of soil erosion in the mountainous areas of southwest China.