Population and industry are closely related to CO2 emissions in Cities. However, few studies have explored the joint influence of population size and industrial structure on CO2 emissions. This paper examined the nonlinear influence of population size and industrial structure on CO2 emissions by using a threshold-STIRPAT model with the latest available data in 2001–2017 from 255 cities in China. Results indicated that the promotion effect of urban population size on CO2 emissions increased in the first two stages and then decreased in the third stage when the industrial structure exceeded the threshold value of 1.22. Meanwhile, the industrial structure had a positive impact on CO2 emissions if the urban population was less than 1.38 million. However, the previous promotional effect became an inhibitory effect when the urban population exceeded 1.38 million. According to the above findings, it is necessary to find a reasonable match between urban population size and industrial structure. Specifically, China should formulate differentiated urban population policies in cities with different industrial structures. In addition, for cities with a population size of more than 1.38 million, adjusting the industrial structure to give priority to the tertiary industry will be an effective way to reduce CO2 emissions.