The slack-based measure (SBM) model was used in this study to calculate the urban green innovation efficiency (GIE) of Chinese 283 cities during 2008–2018, and the night light data from the defense meteorological satellite program/operational linescan system (DMSP/OLS) were used to characterize the economic development level. On the basis, efforts were made to analyze how ecological footprint is affected by urban GIE at varying economic development levels under the Hansen threshold regression model and reveal the mechanism for ecological footprint to receive influence from urban GIE through the mediation effect model. The results show that: (1) The improvement in the urban GIE of the investigated cities during the study period has a negative double threshold in influencing ecological footprint throughout China. However, with higher economic development level, the inhibitory effect gradually weakens, with the elastic coefficient changing from − 0.3046 and − 0.2132 to − 0.1392 at a 1% significant level. (2) The inhibitory effect on ecological footprint from urban GIE is spatially heterogeneous in Chinese cities. In eastern cities other than central and western cities, urban GIE exerts the strongest inhibitory effect on ecological footprint, with the corresponding coefficient being − 0.3972 at a 1% significant level. Moreover, the inhibition in eastern and central regions is strengthened with higher economic development level. Nevertheless, before crossing the second threshold, the inhibitory effect of urban GIE on ecological footprint in western China does not appear, with the coefficient being 0.1899 and 0.1379, respectively, with at a 1% significant level. (3) Industrial structure and energy structure play a mediating role in the effect of urban GIE on ecological footprint. By contrast, population aggregation and infrastructure are important driving forces for the increase of ecological footprint.