Abstract: Soil erosion by water is a global environmental threat, particularly in Mediterranean regions where conventional RUSLE K-factors (Wischmeier & Smith, 1978) often overestimate soil loss due to lack of local calibration. This study develops a locally calibrated K-factor (K_ID) based on the plasticity index (PI) and lithology, with field validation using a custom-built rainfall simulator in the Moroccan Rif. An initial logarithmic regression between K_observed and ln(PI) gave R² = 0.465. After integrating a lithological factor (F_lith_capped) and a final correction factor (0.603587229), the model improved dramatically to R² = 0.9918. K_final_scaled values ranged from 0.0332 (calcareous soils) to 0.2662 (flysch soils), with estimated soil loss ranging from 74 to 600 t/ha/year (mean 39 t/ha/year). Field measurements using the rainfall simulator showed actual soil loss of 10-12 t/ha/year, meaning the developed K-factor reduced the estimation gap by 32% compared to the conventional method (52 t/ha/year). The proposed PI-based K-factor (K_ID) offers a simple, low-cost, and transferable approach for improving RUSLE erosion modeling that can be applied in other Mediterranean basins and data-scarce regions worldwide. Keywords: Soil erosion; RUSLE; Soil erodibility factor (K); Plasticity Index (PI); Rainfall simulation; Éric Roose methodology; Mediterranean Basin; Morocco.