An objective of the soil research program has been to measure forest soil erodibility for erosion prediction. Gathering the data is challenging because samples have to be collected onsite under actual field conditions. These conditions include sites before and after timber harvest, prescribed fire, and wildfire. During prescribed fires and wildfires, a patchwork of low, moderate, and high severity burn areas occurs, and each severity must be measured and characterized to accurately predict erosion. This video shows the field research techniques, which include rainfall simulation on plots from 0.5 to 75 m^2. Plot size depends on the soil property (hydraulic conductivity, interrill erodibility, and rill erodibility) of interest. Natural rainfall on small catchments, with runoff flow measuring devices and sediment traps, is used to validate the erosion prediction models by allowing a comparison of observed runoff and sediment delivery to predicted values.