We propose a method to noninvasively evaluate the permeability of the cell membrane in the rat brain using diffusion magnetic resonance imaging (MRI). Diffusion MRI reflects the intra- and extracellular diffusion coefficients of water and cell membrane permeability. The images were acquired using a 4.7 T MRI system with applications to motion-probing gradients in six directions. Numerical simulations based on the finite-difference method were carried out for estimating diffusion MRI signals with various combinations of membrane permeability and intracellular diffusion coefficient values. We defined an evaluative function as the difference between the signals estimated by simulation and experimentally obtained signals. We found that the combination of membrane permeability and intracellular diffusion coefficient in the simulation corresponding to the minimum value of the evaluative function leads to an estimation of these properties of the rat brain. The estimated intracellular diffusion coefficient and membrane permeability were (1.3±0.1)×10-3 mm2/s and 74±23 μm/s, respectively. Our method is useful for noninvasively estimating the cell membrane permeability of biological tissues, and is easily applicable to human tissues.