The brain can be conceptualized as a complex network. It is believed that the topological structure of the brain network is closely related to the functions of the brain [1]. Therefore, understanding the network structure of the brain is a crucial task in neuroscience. In this report, we propose a first draft of the network architecture of the Drosophila connectome at the mesoscopic scale. The structural network of the Drosophila brain is constructed from a dataset of more than 20,000 single neurons in female brain assembled in the FlyCircuit database (http://www.flycircuit.tw)[2], the most comprehensive database of single-neuron images of the Drosophila brain to date. The nodes of the network represent mesoscopic brain regions called Local Processing Units (LPUs). The weight of the edges connecting each node pair corresponds to the number of neurons innervating the two LPUs reciprocally. The network shows hierarchical structure, pronounced small-world characteristics with high clustering and high global efficiency, and it is composed of six modules corresponding to known functional domains including the sensory modalities (including olfactory, mechano-auditory, and visual), together with the pre-motor and motor centers. Based on the modular structure of the network, we propose two models for the flow of information associated with intuitive and reasoning behaviors, respectively.