This paper uses satellite readings of nitrogen (NO2) air pollution, a byproduct of combustion, to improve the measurement of global economic activity. The proposed approach improves upon night light measures for countries where data manipulation, conflict, or other factors have led to poor national accounts. The paper also shows that existing country rankings of gross domestic product accuracy over the past 15 years are unreliable, even among advanced economies. For example, the paper shows that during COVID, in France, the UK and Spain gross domestic product in 2020 was underreported by 76, 181, and 205 basis points respectively.The methodological contribution extends previous error measurement frameworks which suffer from error-in-variables biases with an objective, data-driven identification strategy exploiting the plausibly orthogonal measurement errors between nitrogen dioxide and night lights, which are measured at different times.