Philosophers of science and epistemologists ask the word 'belief' to cover far too much of the cognitive landscape. When we attend to the epistemic and practical judgments that real scientists make about their theories, we find them quite naturally and effortlessly adopting complex cognitive attitudes that display an extraordinary richness, nuance, and variety. Despite the fact that qualified, partial, and tentative attitudes are a commonplace, our understanding of them remains oversimplified and underdeveloped. The overall aim of the dissertation is to remedy the situation by systematically examining the range of epistemic attitudes and their adoption in scientific settings. The project begins by anchoring the discussion of attitudes in a series of examples from actual science, historical and contemporary, in which it is clear that the adoption of an attitude other than the belief that a theory is true is crucial to an appreciation of that episode. I then develop a framework for thinking about attitudes that helps us to attend to this diversity and complexity at a new level of depth and precision. I construe attitudes as intentional mental states that involve contentful and evaluative responses. The account that I propose, which I call the SA(r)O schema, is constructed around components that should go into any account of attitudes: subject, object, content, and cognitive role. I consider a number of dimensions along which attitudes can vary, explore the epistemic significance of these dimensions, and develop a taxonomy of attitudes important to the scientific enterprise, including assuming as a working hypothesis, conjecturing, speculating, deciding to rely upon, provisionally endorsing, and the like. The project offers an extended discussion of a handful of attitudes, with a focus on the belief-acceptance distinction and pursuitworthiness judgments. I then show how the emerging picture complicates traditional positions on the issue of epistemological voluntarism, casts light on deliberative processes important in scientific practice (such as the nature of theory choice and abductive reasoning), and provides tools for understanding the different ways in which attitudes enter into the debate over scientific realism.