A power is a property in virtue of which its bearer can act or be acted on. This dissertation is about Aristotle’s ontology of causal powers. It highlights three features that differentiate powers from non-powerful properties, to wit: content, the property a power transmits; efficacy, the “oomph” that allows a power to transmit that property; and relationality, the directedness of a power towards its target. It proceeds by way of three case studies, each of which elaborates a puzzle in Aristotle’s metaphysics or natural philosophy, and shows how a better understanding of one of these three features can dissolve that puzzle. The joint activity of a pair of correlative powers is the transmission of some property from the bearer of the active power to the bearer of the passive power. For instance, heating is the activity of a stove’s power to heat, and being heated, the activity of a kettle’s power to be heated. However, heating and being heated aren’t two, numerically distinct process, but two dimensions of the single process whereby the property, heat, is transmitted from stove to kettle. Roughly the content of a power is the property which that power transmits: heat is the content of the power for heating. The content role is played by a form instantiated by the bearer of the active power: the form of heat, for instance, is instantiated by the stove. I examine content by considering Aristotle’s fundamental powers, the elementary qualities. I argue that Aristote is committed to two accounts of the elementary qualities: (i) an account on which the hot (for instance) is a power for heating, and (ii) an account on which the hot is a power for combining like/like. Whereas on the first account, the hot is naturally understood as a principle of transmission, on the latter it is not. Moreover, the duality of Aristotle’s account of the hot threatens its claim to count as one power, let alone an elementary power; this worry is especially pressing given Aristotle’s commitments about power individuation. I argue that the characterization of the hot as a power for heating is fundamental, for it is by heating that the hot combines like/like. I thereby secure the status of the elementary qualities as principles of transmission, and establish their unity. A power is not exhausted by its content. This is clearest in the case of rational powers, like the building art: if having the form of house, the content of the building art, were sufficient for having the art, then every house would also be a builder. The efficacy of a power is that feature which differentiates the builder’s form (which is also a power) from the house’s form (which is not a power). Additionally, while some properties, like heat, are also powers, others, like roundness, are not. In addition to being a positive feature of its bearer, heat is also a cause of heating, while roundness is not similarly a cause of rounding. Thus, efficacy is also that feature which differentiates properties like heat from properties like roundness. The efficacy role, I claim, is played by the activity to which a power is telically ordered: the key difference between the building art and the house’s form is that the former is, in some sense, for the sake of building, while the latter isn’t. I examine efficacy by considering why Aristotelian powers are better-suited to being efficient causes than are Platonic Forms. Aristotle seems to criticize Plato’s view that Forms are efficient causes on the grounds that they are inefficient, but one may worry that if Aristotelian powers are numerically identical to immanent forms, they are no better suited than are Platonic Forms to being efficient causes. I clarify Aristotle’s criticism of Platonic Forms as efficient causes, arguing that Aristotle’s key concern isn’t that Forms are inefficient, but that even if they were efficient, they would fail to explain the effects we actually observe. Thus, the fact that every power is numerically identical to some immanent form is no obstacle its being an efficient cause. I further locate a power’s efficacy in its activity: qua form, a power is indeed inefficient, but qua ordered towards an activity, it is not. Finally, it is a striking feature of powers that they come in correlative pairs. Each power indexes its bearer to another, the bearer of the correlative power: the active power to heat indexes the stove to the class of heatable things, and the passive power to be heated indexes the kettle to the class of heating things. A power exhibits relationality inasmuch as (part of) what it is to have a power is to stand in a possible causal relation to another. The relationality of powers differentiates them from another kind of kinetic principle, i.e. natures. Aristotle contrasts powers and natures by observing that whereas a power is a principle of change in another or in oneself qua other, a nature is a principle of change in oneself qua oneself. I defend a novel account of this distinction. To motivate this account, I first develop a puzzle: the most natural way of understanding the power-nature distinction threatens to misclassify passive powers and natures, for a passive power is also a principle of change located inside the bearer of the principle; however, any modification to the distinction sufficient to preserve the status of passive power as powers compromises the status of elemental natures as natures. I then present my account of the power-nature distinction as a satisfying solution to this puzzle. On my view, insofar as one has a power, one is the bearer of either a form or an activity, but insofar as one has a nature, one is the bearer of both a form and an activity. My view has implications for how we should think about the activities of powers and natures respectively: I suggest that the activity of a power consists in the transfer of a form, while the activity of a nature consists in the process of coming into one’s form.