I defend the conserved quantity theory of causation against two objections: firstly, that to tie the notion of "cause"to conservation laws is impossible, circular or metaphysically counterintuitive ; and secondly, that the conser quantity theory entails an undesired notion of identity through time. My defence makes use of an important meta-philosophical distinction between empirical analysis and conceptual analysis. My claim is that the conserved quantity theory of causation must be understood primarily as an empirical, not a conceptual, analysis of causation.