Rill erosion is the most common form in steep and long hillslopes, but studies about the biochar effects on soil detachment capacity in rills are scarce. This study has simulated soil detachment in rills (D-c) by a laboratory flume on samples collected from forest soils in Northern Iran. The samples were treated with biochars (from rotten fruits of peach and almond, sugarcane bagasse, and straw of wheat, corn, and cotton) or untreated. D-c has been measured under three soil slopes and four water flow rates. The erodibility factor (K-r) and critical shear stress (tau(c)), important parameters for erosion models, have been estimated from interpolations between D-c and the shear stress. Regression equations have also been proposed to predict D-c from stream power or its unit value. The experiments have shown that wheat biochar is the most effective in reducing D-c (-35% compared to the untreated soils). This reduction has been associated with higher organic matter (+65%) and stability of aggregates (> 80%) as well as with lower bulk density (-10%) of soils treated with this biochar. These effects may be due to the highest content of organic carbon, specific surface area, and cation exchange capacity of the substrate. No significant differences in D-c have been measured for the soils treated with the other biochars. K-r was 15%-35% lower compared to the untreated soils, while tau(c) noticeably increased after the treatment with all biochars. The stream power was a more accurate D-c estimator than its unit value (Nash and Sutcliffe's coefficient > 0.90). This study helps land managers and hydrologists select the most effective substrate for soil conservation and estimate rill erosion in forests under similar environmental conditions.