One major focus in epistemology in recent decades has been testimony and our pervasive epistemic dependence on others. Comparatively little has been written on expert testimony, despite our most pressing questions requiring expertise and experts often offering divergent answers. This dissertation addresses the nature and evaluation of expertise. Chapter 1 critiques existing analyses of expertise and proposes this definition: being in a significantly better epistemic position with regard to a domain than the best member of a reference class. Better epistemic position includes having more true beliefs and fewer false beliefs, but it also encompasses other important factors, being multi-dimensional in nature. The next two chapters further explicate these elements. I first lay out what it means for one person to know more than another. This involves a defense of the number of one's true beliefs being part of the measure of knowledge, with their importance as the other component. The many other epistemic desiderata involved in being in a better epistemic position are discussed in the next chapter, derived from our goals for expertise: receiving the truth, having warranted confidence, and the expert's diachronic improvement. The various properties that epistemologists have put forward as essential to knowledge or justification are shown to all contribute to greater expertise in varying ways and degrees. Chapter 4 evaluates indicators, indirect measures that some claim can help us estimate expertise. They include credentials, the number of agreeing experts, biases, track record, and dialectical superiority. However, most of these indicators turn out to be often unreliable and not particularly useful in practice, particularly for laypersons faced with disagreeing experts. Given this situation, the final chapter examines three other approaches to the lay appraisal of expert opinion: prima facie justification, skepticism, and simple trust. None of these are satisfactory, so in order to be in a good epistemic position themselves with regard to expert-related questions, laypersons must directly evaluate the arguments the experts adduce. Experts then shift from being sources of mere judgments to collectors and disseminators of relevant evidence regarding a question.