The aim of this study was to demonstrate the feasibility of the recently introduced Vascular, Extracellular, and Restricted Diffusion for Cytometry in Tumours (VERDICT) framework for imaging prostate cancer with diffusion-weighted magnetic resonance imaging (DW-MRI) within a clinical setting. The VERDICT framework is a noninvasive microstructure imaging technique that combines an in-depth diffusion MRI acquisition with a mathematical model to estimate and map microstructural tissue parameters such as cell size and density and vascular perfusion. In total, 8 patients underwent 3-T MRI using 9 different b values (100-3000 s/mm). All patients were imaged before undergoing biopsy. Experiments with VERDICT analyzed DW-MRI data from patients with histologically confirmed prostate cancer in areas of cancerous and benign peripheral zone tissue. For comparison, we also fitted commonly used diffusion models such as the apparent diffusion coefficient (ADC), the intravoxel incoherent motion (IVIM), and the kurtosis model. We also investigated correlations of ADC and kurtosis with VERDICT parameters to gain some biophysical insight into the various parameter values. Eight patients had prostate cancer in the peripheral zone, with Gleason score 3 + 3 (n = 1), 3 + 4 (n = 6), and 4 + 3 (n = 1). The VERDICT model identified a significant increase in the intracellular and vascular volume fraction estimates in cancerous compared with benign peripheral zone, as well as a significant decrease in the volume of the extracellular-extravascular space (EES) (P = 0.05). This is in agreement with manual segmentation of the biopsies for prostate tissue component analysis, which found proliferation of epithelium, loss of surrounding stroma, and an increase in vasculature. The standard ADC and kurtosis parameters were also significantly different (P = 0.05) between tissue types. There was no significant difference in any of the IVIM parameters (P = 0.11 to 0.29). The VERDICT parametric maps from voxel-by-voxel fitting clearly differentiated cancer from benign regions. Kurtosis and ADC parameters correlated most strongly with VERDICT's intracellular volume fraction but also moderately with the EES and vascular fractions. The VERDICT model distinguished tumor from benign areas, while revealing differences in microstructure descriptors such as cellular, vascular, and EES fractions. The parameters of ADC and kurtosis models also discriminated between cancer and benign regions. However, VERDICT provides more specific information that disentangles the various microstructural features underlying the changes in ADC and kurtosis. These results highlight the clinical potential of the VERDICT framework and motivate the construction of a shorter, clinically viable imaging protocol to enable larger trials leading to widespread translation of the method.