This paper is intended to investigate the statistical significance of the ozone recovery amounts computed with the chemical transport model (CTM) MEZON, given errors in the reaction rates and model initial conditions. The CTM is used to investigate the sensitivity of the ozone to Montreal Protocol and its Amendments (MPA) restrictions on the emission of ozone depleting substances in the atmosphere during 1992-2002. The known Student's t-test is applied for the assessment of the statistical significance of the "saved'' total ozone (TO) by the MPA. The results reveal that the uncertainties in the initial fields do not substantially reduce the statistical significance of the results. Substantial reduction of the statistical significance is caused by the uncertainty in the kinetic constants. The model noise level reaches its highest values (up to 10%) in the polar areas during late winter-early spring. The "saved'' TO also reaches its maximum there and, therefore, it is significant at the 90-95% confidence level.