<p>We sometimes appeal to pure mathematical facts to explain physical phenomena. But it is unclear how a pure mathematical fact could possibly explain something physical. I develop a concrete formulation of this problem informed by the view that explanations of physical phenomena are backed by relations of dependence. I show that prominent analyses of mathematical explanations of physical phenomena fail to provide satisfying solutions to the problem, and I develop a novel analysis that succeeds in their place. I claim that a mathematical explanation of a physical phenomenon works by establishing that its target shares dependencies with a fact that is independently-understood. My analysis dissolves the problem and enjoys many other benefits besides.</p>