Transport generates many externalities, some related to atmospheric pollution. In this paper, we focus on two: greenhouse gases, and local pollution. In the search for optimal transport policies, these two externalities have usually been analysed separately. Here, we study them jointly, in a sequential decision-making model. Our model allows for the irreversibility of the policies undertaken, as well as the possibility of a progressive reduction of uncertainties with the arrival of information. We find that when both sources of externalities are analysed jointly, structural measures enabling private transport requirements to be reduced are identified as being more advantageous economically than technological measures to reduce emissions of pollutants. We illustrate the usefulness of a joint analysis of externalities with two examples: tax measures on cars and housing policy.