Models of evaluation use that appear to underlie most current evaluation practice are examined, and an alternative model is outlined which explicitly addresses the information needs of an innovating manager. In this model, innovation and evaluation are viewed as interlocked processes. An important feature of the user-focused model is that evaluation findings are, at least initially, made available only to the focal users, rather than being distributed widely, as in the traditional model. Further, in conventional evaluation, the designer's central concern is with identifying and obviating threats to the validity of the findings. While these validity concerns remain important in the user-focused model, some degree of relaxation is possible, and a new set of pragmatic concerns that threaten the utility of the study become more serious. In the user-focused model, threats to credibility, user relevance, timeliness for user decisionmaking, feasibility, and acceptable evaluation costs are considered. Aspects of methodology for conducting user-focused evaluations are discussed. Nine references are provided.