A model is presented which can be used to simulate polymer swelling and drug release from tablets using a combination of the Discrete Element Method (DEM) and solving mass transfer over an unstructured grid formed by the DEM elements. This work extends a recently developed single-component swelling and dissolution model by incorporating an extra component (drug) that can be either homogeneously or heterogeneously dispersed within the tablet. Parametric studies were conducted on some of the key parameters which affect drug release, namely the drug distribution, maximum swelling ratio of the polymer and the drug-polymer diffusivity dependence. The drug heterogeneity study showed that burst release (i.e. dose-dumping) increased as drug loading increased since the polymer could not form a gel layer to surround the tablet but the time taken for 99% drug release was not significantly different when compared to the homogeneous case. The maximum swelling ratio study showed that drug release could be limited either by its diffusion through the dense polymer network or the thickness of the gel layer. It was shown that a minimum release time exists where the drug diffusivity and gel layer thickness would give fast drug release. The drug-polymer diffusivity study showed that drug release could become closer to zero-order the more dependent it was on polymer concentration (as the drug would diffuse very slowly until the polymer was very dilute) compared to more Fickian release if the drug's diffusion coefficient was not as dependent on polymer concentration. Simulated drug release was compared against experiments where tablets containing 10% or 60% w/w nicotinamide and HPMC k100LV were dissolved and imaged using a combination of Attenuated Total Reflection Fourier Transform Infrared (ATR-FTIR) spectroscopic imaging and downstream monochromatic UV/Visible detection. The FTIR images and extracted concentration profiles showed that water ingress and drug release were faster for the 60% nicotinamide tablet and that the model could predict drug release using only pure component parameters and the respective mass fractions of drug. (C) 2012 Elsevier B.V. All rights reserved.