A number of climate modeling studies have shown that differences between typical choices for representing ozone can affect climate change projections. Here we investigate potential climate impacts of a specific ozone representation used in simulations of the Hadley Centre Global Environment Model for the Coupled Model Intercomparison Project Phase 5. The method considers ozone changes only in the troposphere and lower stratosphere and prescribes stratospheric ozone elsewhere. For a standard climate sensitivity simulation, we find that this method leads to significantly increased global warming and specific patterns of regional surface warming compared with a fully interactive atmospheric chemistry setup. We explain this mainly by the suppressed part of the stratospheric ozone changes and the associated alteration of the stratospheric water vapor feedback. This combined effect is modulated by simultaneous cirrus cloud changes. We underline the need to understand better how representations of ozone can affect climate modeling results and, in particular, global and regional climate sensitivity estimates.