Educational needs assessments (hereafter NAs) are inattentive to cost considerations and are frequently dominated by elite stakeholder groups. In this article I make a case for adopting a cost-utility approach, illustrating the argument with data generated in a NA of central library services in a Canadian school district. Using survey data from eight stakeholder groups, I found that (1) NAs based on the service preferences of a single stakeholder group can be misleading; (2) service preferences can be integrated into a single set of priorities, even when there are disagreements, by using the stakeholder group as the unit of analysis and assigning weights that privilege input from knowledgeable respondents; and (3) that the ranking of service operations produced by user preferences was not significantly correlated with the ranking produced by integrating preferences with costs. Cost-utility analysis would be more helpful if the utilities represented rigorously determined benefits of the services assessed, as well as stakeholder perceptions of the value of these benefits. Cost-utility analysis in NA will not reach its potential until cost considerations are routinely included in educational program evaluations. (Contains 8 tables.)