In studies of the greenhouse gas implications of convergence by developing countries to the per-capita GNPs of developed countries, considerable discussion has centred on whether purchasing power parity (PPP) or market exchange rates (MER) should be used in measuring per-capita GNPs. We suggest that technology gaps between developing and developed countries should be the starting point for convergence analysis rather than per-capita GNP gaps. We estimate two sets of initial technology gaps, using PPP or MER price assumptions combined with input-output data. In simulating the effects of closing technology gaps (convergence) using a dynamic, multi-country CGE model, we find: (1) the MER/PPP distinction matters. MER-based estimates of initial technology gaps lead to higher estimates of convergence-induced growth in greenhouse-gas-emitting industries in developing countries than do PPP-based estimates. (2) the industry detail in CGE models is valuable. Our simulations show a wide range of convergenceinduced changes in output across industries.