The rate at which limestone dissolves determines the resistance of buildings and monuments to weathering, the efficiency of carbon capture in deep geological reservoirs, and the processes by which soils, rocks, and landscapes form and evolve. However, the normalized rates of mineral dissolution measured in laboratory experiments are often found to be far greater than those measured in field settings. Here, we use atomic force microscopy (AFM) measurements to demonstrate experimentally that the rate of calcite dissolution within micron-size pores at the surface of a limestone sample is much lower than the rate of dissolution in the surrounding calcite surface. In addition, we use numerical simulations to show that this difference cannot be explained using a simple diffusion-surface reaction model. We suggest that the observed heterogeneous reaction rates could instead be due to the elevated density of reactive high curvature features on the polished surface surrounding the pore. These high curvature features can strongly affect local interfacial free energy, making such surfaces more prone to dissolution. As a result, polished surfaces should be more reactive than pore surfaces that have effectively been smoothed during prolonged contact with natural fluids. As rate experiments routinely use polished and powdered samples, the results may help to explain the widely reported discrepancy between lab and field-based dissolution rates. ?? 2013 Elsevier Ltd.