Discussions of the foundations of evolutionary theory -- especially natural selection, fitness, and genetic drift -- are saturated with terms referring to various kinds of chance, stochasticity, randomness, unpredictability, and so forth. This dissertation examines these uses of chance in philosophical and historical perspective. I begin by arguing that, both in the contemporary and historical arenas, the current state of the literature on chance is deeply troubling. Work in the philosophy of biology (i) often conflates various clearly distinct notions of chance, and (ii) often approaches the analysis of chance from the perspective of a debate (on the causal potency of natural selection and genetic drift) that does not in fact profitably engage evolutionary theory. Historically, as well, the most common way of analyzing the development of the use of chance in evolutionary theory does not engage the actual research of historical actors, a point I make by exploring the work of Karl Pearson and W.F.R. Weldon at the turn of the twentieth century. I thus propose a new guiding question for research into the role of chance in evolutionary theory: what is the relationship between our statistical biological theories and the processes in the world those theories aim to describe? I then offer a novel framework for determining the answer to this question, derived from a deeply biologically-informed understanding of fitness, selection, and drift. This view combines core insights from work in philosophy on the propensity interpretation of fitness with cutting-edge biological treatments of population modeling. Chance enters this model at only a single point -- the distribution over the various possible lives that an organism might live -- and this single source can explain the influence of chance throughout fitness, natural selection, and genetic drift. This framework, I claim, constitutes a fruitful way to understand both the foundations of evolutionary theory and the role of chance in those foundations.