Vegetation can effectively prevent soil loss and play an important role in soil and water conservation. Accurate estimation of hydraulic parameters is critical for soil erosion models. Hydraulic data for different levels of vegetation stem cover, slope gradient and flow discharge are quite limited. The objectives of this study were to compare different measured methods of flow velocity, evaluate the influence of vegetation stem cover, slope gradient and discharge on hydraulic parameters and predict the shear stress (τ), stream power (Ω) and unit stream power (ω) and emphasize the significance of hydraulic radius in the calculation of the shear stress (τ), stream power (Ω), and unit stream power (ω). A non-erodible flume bed was used in this study. The discharge ranged from 0.5 × 10−3 to 2.0 × 10−3 m3 s−1, the slope gradient ranged from 8.8% to 25.9%, and an artificial stem cover of approximately 0, 1.25, 2.5, 5, 10, 15, 20, 25 and 30% was used to represent the natural vegetation. The stems were 2 mm in diameter and randomly arranged. The electrolyte pulse method was preferred for measurement of the overland flow velocity. The flow velocity decreased as the stem cover increased, and flow discharge and slope gradient decreased. The reduction in flow velocity was as high as 90% for 30% stem cover based on the experiments. The flow depth increased as the stem cover and flow discharge increased as the slope gradient decreased. The Reynolds and Froude number values decreased with increasing stem cover. The shear stress, stream power and unit stream power were all significantly affected by the stem cover and could all broadly be described using an exponential delay function of the stem cover. The ω was no significantly impacted by hydraulic radius. Thus, the unit stream power is an improved hydraulic parameter for predicting sediment transport capacity.