The purpose of this study is to propose a new measure of nonlinear dissimilarity between two brain regions and to estimate a functional cluster of multichannel EEG. EEG is a summation of electric potentials by recording electrodes placed on the scalp generated from the flow of dendritic currents across the resistance of the extracellular tissue fluids, and it is a reliable indicator of neuronal activity. Many linear and nonlinear methods have been proposed and used for estimating the clinical and psychological researches. Coherence analysis has considered as a useful tool in estimating the functional connectivity. However, this measure only detects the linear properties. Thus, we will provide a new method to provide the nonlinear dissimilarity of EEG between two brain regions. We present the usefulness of this measure showing the results of the numerical and experimental applications. On the other hand, functional clustering provides a useful tool to characterize the joint interactions among many brain regions. It can be defined as a set of elements that are much more strongly interactive among them than with the rest of the system in a specific brain state. The brain regions belonging to the same cluster were all functionally involved while the regions belonging to separate clusters were presumably functionally unrelated in an experimental paradigm or in a group of subjects. This analysis is a useful tool in understanding the global activations pattern of brain in various states. We present the application of functional cluster analysis to multichannel EEGs in various cases including the comparison with other clustering techniques such as PCA and ICA.