Evaluations are designed to provide useful information, often to aid program decisions. Typical evaluation approaches are likely to seriously bias decisions against programs, however, in part because they rely on past data, whereas decisions are related to the future. In addition, assessment of the net differences between benefits and costs is likely to be highly dependent on the timing of the evaluation. Moreover, large programs, such as the Community Anti-Crime Program, are likely to take longer than smaller programs to yield a net excess of benefit over cost. Decision-theoretic approaches to evaluation can be successfully used in decisions at the project level and decisions regarding the refunding of programs. Decision analysis is a proper aid for helping decisionmakers when a particular decision problem is identified. A net utility model of decisionmaking is described and its usefulness is demonstrated. This kind of technique can be used at a fraction of the cost of a typical evaluation. Application of this approach suggests that community anticrime approaches would have become highly cost effective had they not been terminated and that at least some of the anticrime projects achieved a high degree of incorporation into existing community institutions. Footnotes, tables, diagrams, and 36 references are included.