Contemporary environmental changes (i.e., elevated CO2, rising temperature, changing aridity and land use and land cover change) influence C3 and C4 photosynthesis in different ways. This leads to a shift in the photosynthetic advantage of C4 relative to C3 plants, with resulting changes in C4 distribution and global photosynthesis. Current C4 distribution models, however, are inadequate to predict the C4 distribution and the changes in distribution, as they are based on temperature-only empirical relationships and lack observational constraints. In this study, we used a global record of photosynthetic pathways (n > 60,000), satellite observations of herbaceous cover fraction and established optimality models of C3 and C4 photosynthesis to acquire a new observation-constrained estimate of C4 distribution. We found global C4 plants occupied 9.8% - 10.2% of the land surface area during 1992 to 2016. C4 grass stabilized around 7.2% -7.3% as a net effect of negative CO2 impacts and positive impacts from rising temperature and increasing aridity, and C4 crop expanded from 2.5% to 2.9%. We further derived an emergent constraint from an ensemble of dynamic global vegetation models (DGVMs) to estimate C4 photosynthesis from C4 distribution area. We found C4 grass and C4 crop contributed to 9% and 3% of global photosynthesis, respectively, substantially lower than previously recognized (i.e., 25% for all C4) but more in line with the DGVMs ensemble mean (14 ± 13%). The C4 contribution to global photosynthesis has slightly increased in the past 30 years due to increased C4 area, albeit a concurrent decrease in the photosynthesis of C4 crop per unit land area. Our study sheds light on the likely future trajectory of C4 distribution change and improves our understanding of the critical role of C4 in the global carbon cycle.