Since its inception in 1995, more than 200 disputes have been raised under the WTO Dispute Settlement Understanding (DSU). In spite of the obvious numerical success of the DS system of the WTO, in practice several shortcomings call for institutional and/or procedural change. This analysis deals with the economic aspects of the DS system. First, it turns out that the WTO DS system seems to be "biased". The larger and richer trading nations (USA, EU) are the main users of this system, either because of the larger involvement in world trade, or because the LDCs simply lack the legal resources. Second, in taking advantage of recent theoretical explanations of the WTO system in general (trade talks) and the DS system in particular (aberrations from WTO compliance can lead to trade wars) one can theoretically derive the relative robust result concerning the present practice of the WTO DS system: retaliation with tariffs is ineffective, distorts allocation and is difficult to control. This is also demonstrated in an CGE model analysis for the most popular disputes between the EU and the USA: the Hormones, the Bananas and the FSC cases. The major conclusion of our economic evaluation is that the DS system of retaliation should be changed towards a transfer-like retaliation system. (author's abstract) / Series: EI Working Papers / Europainstitut