Objective To investigate the task group’s effectiveness in language evaluation based on the task group's functional Magnetic resonance (fMRI) results’ agreement with the fixation side of the Wada language area. MethodsWe collected 90 patients with intractable epilepsy of 90 patients from December 13, 2018 to January 3, 2020 from the Epilepsy Center of Guangdong 999 Brain Hospital. We used two simple fMRI tasks. Among them, 25 patients completed the Wada experimental examination, and 8 patients completed the electrode implantation and subsequent preoperative language area mapping. Adopt block experimental design, ABBA style presentation, and use AFNI software to process fMRI data, lateralization index calculation, and multiple regression analysis. ResultsfMRI results from 90 patients showed that the results from both the sentence-completion task and the image-naming task were more stable than those from either task. The results were then compared with the results of the “gold standard” Wada test in 25 patients with fMRI-located language dominance in the hemisphere. The results showed that the accuracy of the single task was between 70% and 80%, but the accuracy of the combined results of the two tasks was 93.3%. Conclusions Compared with the results of a single task, the results of multiple fMRI tasks are more stable in the judgment of activation range and language dominance hemisphere. fMRI and Wada language area siding accuracy 93.3%, fMRI task siding valid and replicable.