Abstract The factors governing the amounts of CO, O 2, and O 3 in the martian atmosphere are investigated using a minimally constrained, one-dimensional photochemical model. We find that the incorporation of temperature-dependent CO 2 absorption cross sections leads to an enhancement in the water photolysis rate, increasing the abundance of OH radicals to the point where the model CO abundance is smaller than observed. Good agreement between models and observations of CO, O 2, O 3, and the escape flux of atomic hydrogen can be achieved, using only gas-phase chemistry, by varying the recommended rate constants for the reactions CO + OH and OH + HO 2 within their specified uncertainties. Similar revisions have been suggested to resolve discrepancies between models and observations of the terrestrial mesosphere. The oxygen escape flux plays a key role in the oxygen budget on Mars; as inferred from the observed atomic hydrogen escape, it is much larger than recent calculations of the exospheric escape rate for oxygen. Weathering of the surface may account for the imbalance. Quantification of the escape rates of oxygen and hydrogen from Mars is a worthwhile objective for an upcoming martian upper atmospheric mission. We also consider the possibility that HO x radicals may be catalytically destroyed on dust grains suspended in the atmosphere. Good agreement with the observed CO mixing ratio can be achieved via this mechanism, but the resulting ozone column is much higher than the observed quantity. We feel that there is no need at this time to invoke heterogeneous processes to reconcile models and observations.