In German, this article asserts that present concepts of evaluation are too narrow for curriculum evaluation. Research must differentiate different forms and roles of evaluation, produce a curriculum evaluation model, and resolve problems of values and judgement. The second section discusses the different forms and roles of evaluation: formative and summarative evaluation, micro- and macro evaluation, and intrinsic evaluation as opposed to sequential evaluation. The third section discusses models of curriculum evaluation as developed by 1) R.E. Stake, who defines three areas of evaluation--Hypotheses, Processes and Results; 2) M. Provus who defines a descrepancy model, and 3) D.L. Stufflebeam, who attempts to develop the structure of a design for evaluation. The fourth section discusses the problems of value and judgement in evaluation in relationship to 1) values, priorities, and norms; 2) the programatic and public-political sphere; and, 3) aspects of judgement and decision which are ambiguous in evaluation. The final section lists seven problems which need to be solved in order to develop theories and workable methods of curriculum evaluation. (Author)