This paper empirically examines the separation principle, which asserts that investment decisions are not influenced by dividend decisions. Existing empirical evidence on this proposition is inconclusive. In this paper, we employ causality tests to examine whether investment decisions are, in fact, statistically exogenous with respect to dividend decisions. These tests are undertaken using both firm-specific and aggregate data. The results indicate no causal relationship from dividends to investment, which provides support for the separation principle as an empirical proposition.