Emissions of nitrogen oxides (NOx) lead to formation of ozone, which is an important greenhouse gas. Despite its relevance, little emphasis was previously given on verifying approaches to calculate contributions of individual emissions to ozone and hence to climate change. Basically two methods (perturbation method and tagging method) were used in the past. We demonstrate that both methods are valid and have their area of application, but only tagging calculates contributions of emissions to concentrations, whereas the perturbation method identifies changes in the ozone concentrations due to emission changes. Our results show that the contribution of road traffic emissions to climate change is underestimated by a factor of 5 in the perturbation method. This is caused by non-linear compensating effects from other emission sectors, which are concealed in the perturbation method but disclosed with tagging. Consequently, the effectiveness of mitigation measures for individual sectors (i.e. concentrating on road traffic induced ozone) is only correctly expressed by the tagging method. The perturbation method provides accurately the total impact (i.e. total ozone) of a mitigation measure. However, current approaches, which evaluate the effectiveness of a mitigation measure based on the perturbation approach, do not reflect changes in the chemical state of the atmosphere (i.e. ozone production rates). These largely affect the effectiveness of subsequent measures and hence make the evaluation of the effectiveness of two measures dependent on their chronology of application. We show that also in this regard, the tagging method is better suited to evaluate the effectiveness of a mitigation measure than the perturbation method.