Traditional coronary drug-eluting stents (DES) are made from metal and are coated with a permanent polymer film containing an anti-proliferative drug. Subsequent to stent deployment in a diseased coronary artery, the drug releases into the artery wall and helps prevent restenosis by inhibiting the proliferation of smooth muscle cells. Although this technology has proven to be remarkably successful, there are ongoing concerns that the presence of a polymer in the artery can lead to deleterious medical complications, such as late stent thrombosis. Polymer-free DES may help overcome such shortcomings. However, the absence of a rate-controlling polymer layer makes optimisation of the drug release profile a particular challenge. The use of microporous stent surfaces to modulate the drug release rate is an approach that has recently shown particularly promising clinical results. In this study, we develop a mathematical model to describe drug release from such stents. In particular, we develop a mathematical model to describe drug release from microporous surfaces. The model predicts a two-stage release profile, with a relatively rapid initial release of most of the drug, followed by a slower release of the remaining drug. In the model, the slow release phase is accounted for by an adsorption/desorption mechanism close to the stent surface. The theoretical predictions are compared with experimental release data obtained in our laboratory, and good agreement is found. The valuable insights provided by our model will serve as a useful guide for designing the enhanced polymer-free stents of the future.