The mid-term review of the Agenda 2000 led to the decoupling of direct payments from production. Thus, agricultural production is no longer directly influenced by direct payments and trade distortions have been reduced. The production effects of this policy have been analysed in detail (EUROPEAN COMMISSION 2003a, FRANDSEN et al. 2003, BINFIELD et al. 2004, GOHIN 2006 and KÜPKER et al. 2006). The present study goes further, however, and deals with the question of how decoupling affects the development of farm structures, farm incomes and the capitalisation of direct payments into rental prices in differently-structured regions. The agent-based model AgriPoliS was used to carry out this analysis. Compared to other models, AgriPoliS allows the analysis of structural change with respect to farm growth and farm exit. Therefore, AgriPoliS represents the structure of an agricultural region and simulates the development of farms located in the region. The modelling of space allows the consideration of transport costs and interactions on the land rental market. Farms can also change their production structure by investing in new activities. Furthermore, they change their size by renting or releasing land. Moreover, farms stop production if it is more profitable or if they are illiquid. The objective of this study was to adapt AgriPoliS to six further study regions (Brittany, South East England, the Central Saxonian Loess Region in East Germany, Jönköping and Västerbotten in Sweden, and Vysočina in Czech Republic), given that it has only been applied to the small-structured Hohenlohe Region in Southwest Germany previously. Thereby, it was possible to observe the adjustment reactions of farms located in differently-structured regions. The adaptation of AgriPoliS to additional regions was the basis for the analysis of the mid-term review’s impacts on the farm exit rate, livestock density, farm income and rental prices in the EU-15. The Czech region Vysočina has been analysed separately because of the differences in the political framework between the New Member States and the Old Member States, as well as in the accession process. The objective of comparing the regions in the Old Member States was to find out whether impacts of the mid-term review depend on the initial structure of a region. In addition to the historical model as initially suggested by the EU Commission, the decoupling options chosen by the national governments have also been modelled. Additionally, the bond scheme suggested by SWINBANK and TRANTER (2004) has been analysed. To cover the impacts of decoupling on different farms, individual farm data and data on farms with different legal forms or specialisations have been analysed for the region Hohenlohe and the region Vysočina, respectively. The analyses show that compared to the Agenda 2000, the mid-term review is leading to a reduction in cattle, as well as to a reduction of the farm exit rate. The latter can be explained by the increase in profits due to the stronger market orientation caused by the mid-term review, including the option to receive direct premiums merely for mulching marginal land. For example, many farms ceased beef fattening and are instead maintaining their grassland in GAEC. The regional comparison showed that for all regions, the mid-term review has comparable impacts on livestock production, farm exit rate and farm incomes. However, this does not hold for the rental market. Rental prices increase in regions with a high share of cattle compared to the Agenda 2000 because cattle payments have been redistributed to land. This cannot be observed in South East England and the Central Saxonian Loess Region where cattle are of minor importance. The comparison of the BOND scenario with the mid-term review shows a strong increase in the farm exit rate for the BOND scenario. Simultaneously, there is a run-off of direct payments from the agricultural sector because the exiting farm households still receive the payments. Potential income losses for farms remainning in the sector can be partially or fully compensated, depending on the region. This compensation is possible due to realised economies of scale and due to the reduction of rental prices. Data analysis of model farms remaining in the sector and those exiting showed that for the region Hohenlohe during the policy change, some farms have been misled by their short-term perspective. They stayed in agriculture despite it being more profitable to exit agriculture in the long term. However, such a short-term perspective may exist in reality and could explain the persistence and the very large income disparities in West German agriculture. Analysing the impacts of the introduction of the CAP in the NMS and of the following decoupling of top-ups showed that the impacts of introducing the CAP are significantly stronger than those of decoupling top-ups. The latter are even overlaid by the introduction of the CAP. Furthermore, it was possible to validate the simulation results for the accession scenario with empirical data. The simulation results, as well as the empirical data, show a slight reduction of the farm exit rate due to accession to the EU. The reason for this development is the increase of farm incomes due to the annually increasing direct payments. However, capitalizing payments into land prices is simultaneously increasing. Accordingly, not only do farmers benefit from the increasing direct payments but so do landowners. Decoupling top-ups in 2009 only slightly reduces the farm exit rate. This is probably due to the increasing capitalization of payments into rental prices and the resulting decline in profits. If the New Member States had to decouple top-ups in 2009 as initially planned, and if they had decided for the regional model, a small share of the payments would have been abruptly redistributed among farms. But this did not happen in reality because within the Health Check of the MTR in 2008, decoupling top-ups has been postponed to 2013. Due to this postponement, top-ups will be stepwise decoupled and the regional model will be stepwise introduced until 2013. This gives farms more time to adapt to the new situation. The reason for this development is that top-ups have to be automatically reduced when the sum of top-ups and decoupled area payments equals the total volume of area payments in 2013.