The quantitative estimation of paleoaltitude has become an increasing focus of Earth scientists because surface elevation provides constraints on the geodynamic mechanisms operating in mountain belts, as well as the influence of mountain belt growth on regional and global climate. The general observation of decreasing δ18O and δ 2H values in rainfall as elevation increases has been used in both empirical and theoretical approaches to estimate paleoelevation. These studies rely on the preservation of ancient surface water compositions in authigenic minerals to reconstruct the elevation at the time the minerals were forming. In this review we provide a theory behind the application of stable isotope-based approaches to paleoaltimetry. We apply this theory to test cases using modern precipitation and surface water isotopic compositions to demonstrate that it generally accords well with observations. Examples of the application of paleoaltimetry techniques to Himalaya-Tibet and the Andes are discussed with implications for processes that cause surface uplift.