In situ measurements of stratospheric ozone and laboratory studies of ozone production in electric discharge through oxygen have shown previously that ozone containing heavy isotopes of oxygen (17O, 18O) may be formed preferentially. In order to assess the relevance of the latter experiment to the stratospheric measurements, detailed understanding of the effect of iostopic substitution on the O3 formation reaction O + O2 + M → O3 +M and on the O atom exchange reaction O + O2 → O2 + O is necessary. In this work, an estimate of the effect of isotopic subsitution on the recombination rate is made by use of approximate dynamical theories and statistical mechanics. The results indicate the possibility of isotope effects on the O + O2 recombination rate of the order of several percent at stratospheric temperatures. In general, recombination reactions involving heavy (mass 49, 50) O3 formation are found to be slower than the reaction leading to normal (mass 48) O3 formation. The calculated isotope effects are sufficiently small that the uncertainties in the model input and the approximations in the dynamical theories will probably make the quantitative nature of these results subject to considerable uncertainty. This isotope effect should not be observable in the atmosphere given the precision of the current measurements but may be crucial in the understanding of the laboratory experiments, where observed enhancements are only of the order of several percent. Possible reasons for this discrepancy between the observed enhancement and predicted depletion are presented.