There are a number of approaches open to an information system (IS) development organization wishing to evaluate the potential relevance of an IT product for use in their own organization. This work considers method support for evaluation of CASE tools, as a complex example of an IT product. Our interest is in evaluation for selection, namely the evaluation of products prior to experience of their use in the defined context. Certain weaknesses are apparent in existing evaluation methods proposed in the literature for evaluation for selection. Our primary concern in this paper is to present a new method which, we claim, addresses these weaknesses. We provide a rationale and a detailed description of the new method, which is referred to as 2G. The method specifically addresses the important early phases of an evaluation, during which an evaluation framework is to be established. The method is to be used within the intended usage context for the IT product, so the resulting evaluation framework will be situated rather than general. We identify weaknesses in previously available methods, and discuss the advantages offered by the 2G method. The method has a solid basis in underlying theory, something which has influenced our presentation of it. Equally importantly, our presentation is strongly influenced by our experience with the method in actual usage situations, in small and medium IS development organizations. We claim that 2G is a practical candidate for any IS development organization wishing to undertake evaluations of CASE tools prior to adoption.