The effect on stratospheric temperature of changing ozone is investigated by comparing two 5-member ensembles of 20-year Unified Model transient runs, one with a linear trend in ozone and one without. A significant stratospheric mean temperature trend of -0.17 K/decade is attributed to ozone depletion. It is found that, although increasing the ensemble size to 20 members would have considerable benefits, increasing the ensemble size further would not dramatically improve confidence in the results. The timeslice approach to climate change modeling is found to produce similar temperature trends to the transient approach for this experiment.