This article presents a review of the application of atomic force microscopy (AFM) to crack-tip corrosion during subcritical crack growth in glass. The two principal experimental techniques used in this type of study are (1) the direct observation of crack motion by scanning the tip of a crack during crack growth and (2) the examination of fracture surfaces once the specimen has been fractured in two. The first technique has been used to demonstrate and quantify water condensation at crack tips during subcritical crack growth and is particularly useful at low crack velocities. The second technique has been used to quantify the crack-tip corrosion process and the shape of the crack tip during crack growth. In this article, we discuss experimental results showing that the environment that develops at the tips of freshly fractured glass surfaces in soda lime glass can corrode the glass surfaces near the crack tip. Soda lime silicate glass contains mobile alkali ions that will exchange with hydronium ions in solution at the crack tip, forming a highly basic solution that is corrosive to glass. Experimental evidence for such corrosion has been obtained by the atomic force microscope, which demonstrates a displacement of the two fracture surfaces near the crack tip that can be as much as 20 nm, depending on how long the crack is held open at the fatigue limit. Despite the corrosion and displacement of the crack surfaces, the crack tip itself appears to remain sharp, suggesting that the fatigue limit in soda lime silicate glass is not due to crack-tip blunting. Most likely, the fatigue limit is a consequence of ion exchange at the crack tip, in which hydronium ions in the crack-tip solution exchange with sodium ions in the glass. As hydronium ions are larger than sodium ions, this exchange process leaves a compressive stress within the fresh fracture surface of the glass that resists crack motion and results in a stress-corrosion fatigue limit, as first proposed by Bunker and Michalske. In agreement with this mechanism, no fatigue limit is observed for silica glass, which also exhibits no ion exchange. As the crack-tip solution in silica glass is only mildly acidic, pH ≈ 5, corrosion does not occur at crack tips of this glass as supported by the observation that no crack-tip displacements are observed in silica glass by AFM. As the proposed ion exchange mechanism used to explain the stress corrosion limit in glass is at variance with the belief that the fatigue limit in glass is the result of crack-tip blunting, we discuss the possibility of plastic deformation at crack tips in glass and conclude that the available experimental data does not support such a model. At the present time, chemical reaction based crack growth theories are most consistent with the body of crack growth data that is available on glass and are probably the best explanation for the phenomenon.