Abstract Tropospheric ozone (O3) is harmful to plant productivity and negatively impacts crop yields. O3 concentrations are projected to decrease globally in the optimistic Representative Concentration Pathway of 2.6 W m–2 (RCP2.6) but increase globally following the high-emission scenario under the RCP8.5, with substantial implications for global food security. The damaging effect of O3 on future crop yield is affected by CO2 fertilization and climate change, and their interactions for RCP scenarios have yet to be quantified. In this study, we used the Joint UK Land Environment Simulator modified to include crops (JULES-crop) to quantify the impacts, and relative importance of present-day and future O3, CO2 concentration and meteorology on crop production at the regional scale until 2100 following RCP2.6 and RCP8.5 scenarios. We focus on eight major crop-producing regions that cover the production of wheat, soybean, maize, and rice. Our results show that CO2 alone has the largest effect on regional yields, followed by climate and O3. However, the CO2 fertilization effect is offset by the negative impact of tropospheric O3 in regions with high O3 concentrations, such as South Asia and China. Simulated crop yields in 2050 were compared with Food and Agriculture Organisation (FAO) statistics to investigate the differences between a socioeconomic and a biophysical process-based approach. Results showed that FAO estimates are closer to our JULES-crop RCP8.5 scenario. This study demonstrates that air pollution could be the biggest threat to future food production and highlights an urgent policy need to mitigate the threat of climate change and O3 pollution on food security.