Heavily based on the developments in professional organisations working with competences is more and more subject in the debate about the (re)design of vocational education and learning. According to Velde (1999) the concept of competence-based education (CBE) can facilitate learning in a society characterised by rapid changes and complexity, like today's society. In education, based on competences the content is no longer starting point for curriculum development, but the competences that are important for working in practice, are. CBE is a dominant trend in the Netherlands, UK, United States and Australia (Velde, 1999; Descy and Thessaring, 2001) in VET because of the expected decrease of the gap between education and labour market (Biemans et al., 2004). The implementation of CBE should be seen as a system innovation. In several practice oriented projects researchers as well as people from practice are working on designing competence profiles, integrating courses, bringing the practice closer to the school, and designing competence-based reward systems. Although the concept of CBE is very popular there is not a clear conceptual framework of what CBE actually comprises. Several authors did make a start with defining CBE (Klarus, 2004; Mulder, 2004; Vrieze and Braam, 2004; Onstenk et al., 2004, Wesselink et al., 2004); there is still a need for a thorough conceptual framework including an operationalisation. Central question in this contribution is: 'Which principles characterise competence-based vocational education?' Starting point of this study was a preliminary list with principles for CBE formulated by Mulder (2004). By means of a Delphi study this list is developed further. The choice to use a Delphi study was made because several experts in the field of CBE can give their judgements on the principles of CBE through a set of sequential questionnaires interspersed with summarised information and feedback of opinions derived from earlier responses (Delbecq et al., 1975). Besides that making use of different experts enlarges the chance that people working with CBE accept and support the conceptual framework. A group of 15 experts joined the Delphi study. The participants were selected on the base of their expertise in the field of CBE and come from eight different (research) institutes. The Delphi study can be considered as completed when 75% of the participants agree with the principles and the operationalisation of the principles.This Delphi study consisted of three rounds. The first round was a meeting. During this meeting the first set of principles was discussed. The results of this discussion were processed and sent to the participants. In the last two rounds the participants were asked to complete a questionnaire. In this questionnaire they could mark to what extent they agree with the principles and the operationalisation of the principles and they could also include some comments. The reactions and comments of the participants were processed. The result of the Delphi study is a conceptual framework for CBE in the form of a matrix. The matrix consists of seven principles. These principles are operationalised in indicators and these indicators lead to different phases per principle. Per principle a limited number of indicators is formulated and these indicators are the base for the distinguished phases. Per principle four phases are defined. These four phases represent the development stages of CBE. The development of the matrix is heavily based on theory and research activities. The principles form the theoretical framework. The indicators in the matrix form the operationalisation of the principles. During the study it appeared to be necessary to define two or more indicators per principle. In theoretical sense this means there is a kind of cohesion between these indicators. It is still a question to what extent these indicators really show cohesion in practice. Using the matrix in practice has to show to what extent these indicators really show cohesion. This question is central in the continuing of this study.LiteratureBiemans, H., L. Nieuwenhuis, R. Poell, M. Mulder en R.Wesselink (accepted for publication). Competence based VET in the Netherlands: backgrounds and pitfalls. Journal for vocational education and training. Delbecq, A.L., A. van der Ven & D. Gustafson. (1975). Group techniques for Programme Planning: A guide to nominal group and Delphi processes. Glenview, Ill: Scott, Foresman and Company. Descy, P.& Tessaring, M. (2000). Training and learning for competence. Thessaloniki: Cedefop.Klarus, R. (2004). Omdat het nog beter kan. VELON Tijdschrift voor lerarenopleiders, 25 (4), pp. 18 - 28.Mulder, M. (2004). Educatie, competentie en prestatie. Over opleiding en ontwikkeling in het agro- foodcomplex. Oratie: Wageningen UR. Onstenk, J., E. de Bruijn, J. van den Berg (2004). Een integraal concept van competentiegericht leren en opleiden. Cinop: Den Bosch. Velde, C. (1999). An alternative conception of competence: implication for vocational education. Journal of Vocational Education and Training. London: Triangle.Vrieze, G. & H. Braam (2004). CGO-wijzer. Nijmegen: ITS. Wesselink, R., T. Lans, M. Mulder & H.J. A. Biemans (2004, submitted for publication). Evaluating design principles for competence-based education. Learning and Instruction.