The development of real income series that are comparable both over time and across countries significantly expanded the possibilities for research in quantitative economic history and long-run economic growth.This dissertation was completed in 2000 in the Department of Economics at the University of Miami (FL) under the supervision of Professor John Devereux. Professor Luis Locay also contributed significantly to this work's progress. This dissertation re-examines the long-run GDP estimates that span over 100 years (henceforth, long-span GDP estimates). While these estimates are used extensively in the literature, their reliability has received limited attention.For example, this data are central to the large body of work on catch-up and convergence. The seminal papers in this area are Abramovitz, “Catching Up”; and Baumol, “Productivity Growth.” Elsewhere in the growth literature they have been used to test competing growth theories in papers such as Kocherlakota and Yi, “Is There Endogenous Long Run Growth.” My findings are first, that there are persistent discrepancies between the long-span GDP estimates and alternative benchmark estimates of comparative GDP. Second, the long-span GDP estimates may result in the mismeasurement of economic performance for individual countries and the forces of catch-up and convergence across countries.