Fuzzy Cognitive Maps (FCMs) have been modelled and designed to reflect expert knowledge for yield forecasting of corn. The developed FCM model comprises nodes linked by directed edges, where the nodes represent the main variables that influence corn yields, and the directed edges indicate the weighted relationships between the yield indicators and the corn yield value. The most effective predictors in the study were those that reflected the influence of climatic factors and soil moisture, such as FPAR, LST and SSM. They provided a relative forecast error of 5.1–10.0%. The indicators of vegetation condition, including dry matter accumulation (GPP), leaf area index (LAI) and vegetation indices MSAVI and NDVI, provided a moderately high forecast error of 10.1–15.0%. Nevertheless, achieving such accuracy is acceptable and standard for the production environment.