Even though several national testing firms have developed measures to evaluate the effectiveness of baccalaureate education, there continues to be a general reluctance on the part of faculty in colleges and universities to accept these measures as criteria on which to evaluate educational programs. Some of the resistance appears to lie in the lack of validity of commercially prepared tests as measures of their competencies (often 50 to 100 in number), the amount of cost and effort to evaluate these competencies with the kinds of performance tests they seem to prefer, and questionable psychometric rigor of locally made performance tests. A series of three prototypical real-life problem solving exercises and rating criteria were developed to address these limitations. A theory of human performance provides a rationale for the development of the generic skills test. The stimuli are in the form of real-life scenarios, and require short-answer essay responses. Rating scales use descriptions of high, medium and low performance attributes to evaluate responses. The results of a field test indicate that the difficulty level of exercises appears acceptable, the problem-solving constructs are independent and meaningful, and that the interrater and alpha reliabilities are sufficiently high to warrant its use as a program evaluation instrument. The contribution of basic intellectual factors, maturation and educational experiences to generic performance should be ascertained through further research. (Author)