The formation of ozone by the termolecular association of atomic and molecular oxygen is discussed. On taking the relevant symmetry numbers into account it is inferred that there is an enhancement that depends on &tgr;(R)/&tgr;(D) where &tgr;(R) is the energy randomization time and &tgr;(D) is the lifetime of the activated complex. To explain the observed enhancement &tgr;(R)/&tgr;(D) would have to be about a third which is within the rather wide range that the uncertainties in &tgr;(R) and &tgr;(D) allow. ¿ American Geophysical Union 1988