Soil erosion is a worldwide threat that results in soil degradation, agriculture abandonment, and crop yield reduction. There is a need to find methods to survey soil erosion rates in order to improve and develop sustainable land planning. The present study utilizes new approaches based on the fuzzy set both in designing the problem (through the fuzzy decision making trial and evaluation laboratory) and in prioritizing the effective factors to mitigate soil erosion (using a fuzzy analytical network process, FANP). This study is first to apply these methods to soil erosion. A set of geo‐environmental factors influencing soil erosion was characterized to evaluate the potential risk of soil erosion in the Nor‐Rood watershed in Iran. The layers of information were developed using expert knowledge, and a network structure was designed by the fuzzy decision making trial and evaluation laboratory method. Then, the weights of layers were calculated by the FANP method by considering the internal and external interaction between factors. The erosion susceptibility map was produced by combining layers based on their weights in a geographic information system platform and was validated using erosion occurrences recorded in field surveys. Results revealed that FANP model accuracy is high (83.4% accuracy) for the study area. We found that vegetation, drainage density, land use, and soil erodibility are the key parameters to explain the soil erosion rates. The soil erosion risk map developed by the FANP method provides useful information for sustainable planning and risk mitigation and can be used in a data‐poor environment.