After summarising briefly the national development gap of the CEECs relative to the EU (section 2), a more comprehensive view is taken on the CEECs' regions (section 3). Section 4 presents some theoretical considerations on whether priority should be given to national or to regional development policy. Section 5 discusses several implications for EU structural assistance. Section 6 summarises and concludes.II/039/97-ENSummary and conclusions The gap of national development compared to the EU15 average income as measured by income per capita in 1993 was considerable, the 10 associated CEECs' average being at 29% of EU15 average.Regional policies and the regions in CEECs are characterised by a missing level of regional administration and statistics. Regional policy in the socialist era was essentially the outcome of national sectoral plans and can broadly be categorised into two country groups: A first group of countries, including Bulgaria, the German Democratic Republic, Hungary and Poland, did not undertake major efforts to reduce inter-regional disparities and allocated industries - although with several exceptions - according to traditional locations. A second group of countries, among which Czechoslovakia, Romania and Slovenia, tried to allocate industries to all regions of the country, in this way creating intra-regional differences and locations that have problems to survive in a competitive environment.Although economic transition caused polarisation effects, an implemented regional policy can hardly be found in any of these countries due to disputes over ministerial responsibilities, the need to reform territorial administrative structures and the lack of financial resources. A major regional orientation can only be found in some countries with active labour market policies for problem regions as well as in the context of programmes on cross-border co-operation. Describing regional disparities in the CEECs during the transformation process is limited by insufficient regional data. Unemployment rates in Romania and the Czech Republic vary only little across the regions, while there are major differences within Poland, Slovakia, Hungary, Bulgaria and Albania. Classifying the regions in the four Visegrad countries according to sectoral employment shares in 1991 above the national average suggests a typology of regions as mono-structural problem regions, lagging regions with a development potential and growth poles. This typology of regions is essentially in line with the ones frequently applied in western countries and gives a first idea of the future objectives of regional policies in the CEECs: diversification of the economic structure in mono-structural problem regions, development of the locational advantages in regions with a development potential and removal of bottlenecks in the growth poles. Given the financial constraints of implementing EU structural assistance in CEECs, one of the most crucial decisions will be on the regional allocation of funds, in particular whether to promote primarily national or regional development because a trade-off between national and regional development policy could force governments to decide whether to prioritise either national development or regional equality. The institutional setting in the CEECs' regions, i.e. the lack of regional administration and statistics, as well as subsidiarity considerations suggest not to have structural assistance carried out directly on a regional level.Implications for EU structural assistance bringing together PHARE and Structural Funds assistance were discussed following the four operating principles that were introduced for the reform of the Structural Funds in 1988: concentration, partnership, programming and additionality. In respect to concentration, no new objective for countries in economic transformation should be introduced because the differences in regional problems are of a quantitative dimension rather than of a qualitative one. The basic problem in applying the principle of partnership would be a lack of administrative and social partners at the regional level, but could be an important leverage for democratisation and institution-building at the national level. While programming should in general be no problem at the national level, there might be insufficient information to programme the regional allocation of funds. In the course of enlargement, an option for reform could be to give Member States access either to a Cohesion Fund-like scheme or to a Structural Funds-like scheme in order to allow for a better targeting of projects and instruments. The Cohesion Fund-like scheme would provide for the real convergence of poorer Member States towards the Union, based on a national development programme, while the Structural Funds-like scheme would provide for social and regional cohesion in the wealthier Member States. There might be some specific problems for CEECs to comply with the additionality principle and to avoid excessive deficits at the same time. The main reason is that the budgets of the CEECs still face some risks for revenues and expenditures. Therefore, maintaining structural expenditure at a stable level without allowing for temporary public deficits will be difficult for CEECs.To sum up, several arguments suggest not to target EU structural assistance exclusively at the reduction of regional disparities within CEECs. Doing so would not only face problems due to missing administrations and statistics at the regional level, but might also be at the expense of national economic growth and extend the process of catching-up. A Cohesion Fund-like scheme could be developed into an adequate instrument for assisting the real convergence of cohesion countries towards the Union.