Conventional neuroimaging methods primarily focus on functional localization, through which specific cognitive functions are localized to specific brain regions. However, fully understanding the human brain function requires characterization of functional integration within and among the functionally specialized regions in addition to functional localization. Functional connectivity and effective connectivity analyses have been developed to investigate functional integration in human brain. Several approaches for modeling functional connectivity and effective connectivity, including the time-series correlation, psychophysiological interaction (PPI), structural equation modeling (SEM), dynamic casual modeling (DCM), and diffusion tensor imaging (DTI) are reviewed. The applications of functional connectivity analysis to the studies of object representation, motor coordinate, language, and autism are demonstrated. Functional connectivity study will highly enrich our knowledge about the dynamic integration in the human brain.