Some philosophers of science, such as Harvey Brown (2005) and Margaret Morrison (2000), argue that in physics only low-level theories are explanatory. Low-level theories are those theories which refer to natural laws and describe the underlying physical mechanisms and nature of matter to explain and describe phenomena. Such theories are contrasted with high-level theories which do not make a commitment to these underlying features but refer to high-level regularities, assumptions and empirical observations about the behavior of objects. Using special relativity as a case study, I show that high-level theories are explanatory. I apply and then later adapt Kitcher’s unificationist account (1981 and 1989) to this specific case in order to show that contra Brown and Morrison high-level theories are explanatory. In chapter 2, I show that STR is an example of non-reductive high-level theories in physics. In chapter 3, I apply Kitcher’s unificationist account to the derivation of length contraction and Fizeau’s experimental result provided by the principle interpretation. In chapter 4, I consider Kitcher’s criteria for determining the degree of unification of a theory. I apply his criteria to my case study in order to show that the principle interpretation is more unifying than the dynamical theories of Lorentz and Poincaré, and Brown. I conclude that because the principle interpretation is more unifying than its competitors it is explanatory under Kitcher’s account. In chapter 5, I review and critique Michel Janssen’s defense of the explanatory value of STR (2002 and 2009), in order to show that my characterization of STR using Kitcher’s unificationist account is the best option available. In chapter 6, I move to show that Kitcher’s unificationist account, or at the very least a modification of it, is defensible. To do so, I consider one of the most prominent objections to his account (Woodward 2003). I then modify Kitcher’s account by incorporating dependency relations. I conclude that special relativity is explanatory and that my account of explanation can capture both causal and noncausal explanations in science.