Carbon emissions and population aging have risen as two major challenges of the sustainable development of human society. This study explores the nonlinear effects of population aging on carbon emission by developing a panel threshold regression (PTR) model, using data from 2002 to 2012 of 137 countries or regions. In the PTR model, carbon emission is the explained variable, whereas population aging is the threshold variable; industrial structure and urbanization are the explanatory variables; and level of economic development, trade freedom, population size, and financial level are the four control variables. With the increase in population aging, the correlation between industrial structure and carbon emissions in the high-income, upper-middle-income, low-income groups are positive, negative, and inverted “U” shaped, respectively. Moreover, with the increase in population aging, the correlations between urbanization and carbon emissions in the high-income group is inverted “U” shaped, whereas the correlation in upper-middle-, lower-middle-, and low-income- groups is nonlinear and positive. Display omitted •The effects of different aging lever on carbon emissions are explored.•The Panel Threshold Regression is devloped with the panel data of 137 countries.•Industrial structure and urbanization contribute carbon emissions with aging as the threshold.•Different income groups have different effects.