Rather than automatically presuming explicit conditions exist when designing an evaluation to fit the summative or formative mold, evaluators should think of an evaluation design as fitting between endpoints on an evaluation process continuum. Evaluators can blend techniques from both the formative and summative evaluation, matching actual program conditions, "sliding" along the continuum to meet a primarily formative or summative goal. Using a case study, this paper explores ways in which formative and summative techniques can serve such complementary roles in a single evaluation process. The evaluation of the National Commodity Processing System, a pilot program to reduce government food surplus stocks, illustrates how decision rules arising from presumed, but untested, summative conditions unnecessarily narrowed the research scope and the final product given to decision makers. This left little room for the depth and detail that qualitative data could offer. The continuum evaluation took a much broader approach to fit differing needs. Concepts, definitions, and models should not be ordered in advance, but selected to give balance to achieve the purpose at hand. Evaluating from the standpoint of a constantly creative continuum is one alternative approach to achieve that balance. (PN)