Although the importance of systematic evaluation of educational policy and practice is well established, various problems confound the intention of evaluating regional or national activity with regard to computer-related activities in education. At least two of these problems relate to a general conception of the appropriate entry points for evaluative activity, while at least two other problems relate specifically to computer-related policy and practice. At least one additional layer of difficulty applies when the evaluation of such activity involves an external evaluator working internationally, particularly in the less developed countries. Each of these classes of impediments is briefly analyzed and a general model for evaluating regional and national-level computer-related educational activity is discussed with respect to the classes of impediments and illustrated in the context of an ongoing external evaluation of a national educational software development project in the Netherlands. This activity--titled the "POCO Project" after its Dutch name meaning software development for computers in education--has the organizational goal of developing a first set of 18 software packages in 18 months. It focuses on developing educational software that can be used by teachers in a meaningful way during their regular teaching activities and with such frequency as to strengthen the teachers' perceptions that using such packages is an effective and efficient response to an educational need. The particular advantages of an outside evaluator are emphasized in this description of the POCO Project. (Author/SLD)