The availability of data from different sources or modalities had empowered the capabilities for analysis of different complex phenomena. However, there are several inherent issues to develop a joint analysis of several modalities such as time and spatial synchronization of the data. I this paper, we propose a method for the joint analysis of functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) recordings measured simultaneously. The method considers fusion of features extracted from the data and fusion of the classification results from several methods. Several single classifiers were implemented, which results were fused using average fusion and alpha integration, an advanced fusion method. This later ensures optimality from the point of view of least mean-square error. The results show the capabilities of the fusion to improve the results of single classifiers and possible clinical meaning of the results.