Readout-segmented echo planar imaging (rs-EPI) is a form of multi-shot EPI. rs-EPI is affected less by susceptibility artifacts than single-shot EPI (ss-EPI), which is widely used for diffusion-weighted imaging, so rs-EPI is expected to produce less image distortion. In this study, we compared rs-EPI and conventional ss-EPI of the temporal bone region, which contains abundant amounts of air and frequently exhibits changes in magnetic susceptibility. In addition, we used a phantom to determine the optimum rs-EPI acquisition conditions for clinical use and investigated the clinical utility of rs-EPI in 20 patients (8 men, 12 women, mean age, 54.3 ± 16.7-years-old) with cholesteatoma (mean apparent diffusion coefficient on ss-EPI, 0.88 × 10(-3) ± 0.18 mm(2)/s). The images of the temporal bone region produced using rs-EPI exhibited less distortion than those obtained with ss-EPI (P < 0.05).