Book excerpt: This thesis studies how rocks evolve due to the coupled effects of flow and chemical reaction. The study was motivated by various experimental observations, both in igneous and sedimentary rocks. In the first part of this thesis, growth of microscopic, pore-scale, features in sedimentary rocks is theoretically investigated. It is found, in agreement with experiments, that statistical properties of pore-grain interfaces mirror growth conditions. The shapes of pore-grain intrefaces both influence and are influenced by large-scale transport properties of the rock. The second part of this thesis employs analytical methods to study flow patterns in melt upwelling beneath mid-ocean ridges. It is shown that high permeability channels spontaneously form, allowing for efficient extraction of melt from the system. This result may aid in understanding existing geochemical and geological observations. In the third part of this thesis, I present a new 3D computer model that simulates flow and reaction through a porous matrix. The model is used to study and compare the different characteristics of dissolution and deposition, and to simulate different settings for melt upwelling in the mantle.