It is a truism in the sociology of science that scientific knowledge bears the imprint of particular perspectives, interests, and values. In social science, it is especially common to find that research serves the needs of managers and policymakers better than it serves the needs of front-line workers. This paper analyzes the traces of that tendency in police research. By examining three features of front-line police work (the need to improve programs rather than assess them, the need to attend to an enormous number of situational details, and the need to cope with ambiguous and contradictory goals), I argue that common approaches to police research address managerial and policy concerns better than line officer concerns. To help rectify this imbalance, I discuss three variations on an alternative research strategy that deserves more emphasis in policy-oriented police research*one grounded in concrete case study description and analysis that often eschews causal generalizations.It is a truism in the sociology of science that all knowledge bears the imprint of particular perspectives, interests, and values, so that any particular scientific program serves the priorities of some groups rather than others. From a normative viewpoint it is not even clear that this influence is always nefarious. Aristotelians once valorized a disinterested approach to scientific inquiry, but the idea that knowledge should strive to advance specific human interests is at least as old as Francis Bacon, who insisted that 'it is by witness of works, rather than by logic or even observation, that truth is revealed and established.' As Bacon himself went on to note: 'It follows that the improvement of man's mind and the improvement of his lot are one and the same thing' (Farrington 1948: 93). From this perspective, scientists' views about what it means to 'improve man's lot'*and about whose lot in particular they should try to improve*unavoidably influence the scientific agenda.To claim a connection between truth and (non-cognitive) values is faintly scandalous for the obvious reasons, but at the general level I have invoked this claim it is impossible to deny. At minimum, no one doubts that the questions scientists ask are legitimately shaped by human interests, so which truths they discover (if not the very fact that those truths are true) depends on the interests to which they and those who influence their work are committed (e.g., Taylor 1985). That conclusion is so undeniable that the distinguished and decidedly non-radical philosopher of science William Newton-Smith (1984) dubbed it 'Boring Interest Thesis 1', and it is amply illustrated by agricultural and health research that serves first-world needs more effectively than third-world needs (e.g., by developing cures for diseases associated