Many factors and processes are expected to be altered in the troposphere under ongoing climate change. This is likely to affect the regime of concentrations of tropospheric species. In the case of ozone, the situation is somehow uncertain, because the expected changes are often contradictory in regards of its production and destruction. To project its future concentrations, regional chemical-transport models can be used, but they usually exhibit systematic errors which can originate from many different sources. There are statistical tools that can be used to compensate for these shortcomings, but they are often not designed for the fine resolution simulations typical for regional scale models. Our study applies an innovative method of statistical processing to projections of ozone concentrations performed by WRF-Chem and CAMx models for two future periods under the RCP4.5 and RCP8.5 scenarios. In the area of the Czech Republic, each projection expects a flattening of the annual cycle of ozone concentrations at different rates, yielding higher than present concentrations in winter and lower in summer. The results also project complex changes in the seasonal spatial variability in 2026–2035 under both scenarios, but in 2046–2055 an overall decrease in concentrations under RCP4.5 and an increase under RCP8.5.