After the discovery of the Antarctic ozone hole by Farman et al. (1985) the need for a complete understanding of stratospheric ozone chemistry and related dynamics became apparent. Since the early 1990ties we concentrated our research on following questions: 1. Do similar ozone losses appear in the Arctic stratosphere? 2. How much ozone becomes depleted? 3. Is our understanding of the underlying chemical and dynamical processes correct? Only with a complete understanding the future of the ozone layer can be predicted.To answer the first two questions we developed a Lagrangian method, the so called Match method, to detect and to quantify ozone losses. By means of hundreds ozonesondes launched in near-real time coordination at several stations in the polar and sub-polar region during a winter season we were able to show that processes similar to those leading to the Antarctic ozone hole occur in the Arctic, too, and that the amount considerably varies from winter to winter. However, we found that the amount of ozone losses in winter with great losses increased.It turned out that the Match data set was very well suited for comparisons with model results. First comparisons showed a significant underestimation of the ozone losses by state-of-the-art models. After more than ten Arctic Match campaigns we therefore performed two Antarctic campaigns to enlarge our experimental data base qualitatively. One took place in 2003 and the other one more recently in 2007 within the frame of the IPY project ORACLE-O3. In the mean time the models have been improved and explain experimental data quite well. However, recent new laboratory measurements of a fundamental constant in the ozone loss chemistry cast doubt on our general understanding of the corresponding processes. We will report about our results with respect to the current state of the ozone research.