Diffusion MRI is a technique that can probe direction-dependent diffusivity of water molecules to reflect, on a statistical basis, the displacement distribution of the water molecules present within a MRI voxel. The observation of this displacement distribution may thus provide unique clues to the structure and geometric organization of tissues. The pioneering work on tissue microstructure took off with the introduction of advanced Q-space measurement. From which, 3D measurement of the diffusion-driven displacement of water molecules could be fully derived. Nonetheless, due to the limitation of imaging hardware and scan time for human subjects, plenty of models have been proposed for mapping neural orientations and tissue heterogeneity. The most advanced works for neural orientations mapping are diffusion tensor imaging (DTI), high angular resolution imaging (HARDI), and diffusion spectrum imaging (DSI). In order to map tissue microstructure, bi-exponential model, diffusion kurtosis imaging (DKI), composite hindered and restricted model of diffusion (CHARMED), and neurite orientation dispersion and density imaging (NODDI) model were proposed. Here, I will review the principle of diffusion MRI, the progress to these methods, and their potential applications in neuroscience.