Windstorms in Europe occur mainly in autumn and winter. They are typically associated with the passage of intense extratropical cyclones and/or their frontal systems. During the last 20 years, the probably most damaging storms were the 1987 October storm in the UK and France (Burt and Mansfield 1988), the December 1999 Denmark storm (Nielsen and Sass 2003) the two successive storms in December 1999 over central Europe (Lothar and Martin, Ulbrich et al. 2001) that led to enormous damage in France, southern Germany and Switzerland and Kyrill (January 2007) that caused damages in excess of 2 Billion Euro in Germany alone and brought the German railway system to a total standstill for the first time in history. With the exception of Kyrill these examples of extreme European winter storms were poorly forecast. Kyrill was very well forecast with more than 24 hours lead time so that timely warnings could be delivered to emergency managers and the general public. However, even in this case the current operational forecast models did not completely provide the detailed structure of Kyrill necessary for the efficient guidance of local emergency response staff. All of these storms were associated with particularly strong diabatic processes and intense precipitation and were characterized by a different dynamical evolution. The storms over the UK and Denmark were interpreted as a consequence of favourable interaction between upper and lower-level potential vorticity anomalies (Hoskins and Berrisford 1988; Nielsen and Sass 2003). In contrast, the hypothesis for the explosive development of the “Lothar” storm was a bottom-up development induced by a diabatic Rossby wave below a straight and intense upper-level jet without a notable precursor disturbance at the tropopause level (Wernli et al. 2002). This subtle difference in the basic dynamical evolution is interesting and points to the existence of different mechanisms that lead to explosive cyclone development, and to potential forecast failures. The storms described above highlight also the challenge involved in the prediction of such systems, since their evolution depends crucially on mesoscale structures over the generally data sparse oceans, both at the level of the tropopause (wind, temperature) and in the lowest troposphere (also humidity).