Existing methods for determining paleoelevation are primarily limited by (1) large errors (±450 m), (2) a reliance on incorrect assumptions that lapse rates in terrestrial temperature decrease with altitude in a globally predictable manner, and/or (3) are inherently climate dependent. Here I present a novel paleoelevation tool, based on a predictable, globally conserved decrease in CO2 partial pressure (pCO2) with altitude, as indicated by increased stomatal frequency of plant leaves. The approach was validated using historical populations of black oak (Quercus kelloggii). These analyses demonstrate highly significant inverse relationships between stomatal frequency and pCO2 (r2 > 0.73), independent of ecological or local climatic variability. As such, this is the first paleobotanical method to be globally applicable and independent of long-term Cenozoic climate change. Further, tests on modern leaves of known elevations indicate that species-specific application to the fossil record of Q. kelloggii (= Q. pseudolyrata) will yield paleoelevation estimates within average errors of ∼±300 m, representing a significant improvement in accuracy over the majority of existing methods.