Household financial market participation affects asset prices and household welfare. Yet, our understanding of the participation decision is limited. Using an instrumental variables strategy and dataset new to this literature, we provide the first precise, causal estimates of the effects of education on financial market participation. We find a large effect, even con trolling for income. Examining mechanisms, we demonstrate that cognitive ability increases participation; however, and in contrast to previous research, financial literacy education does not affect decisions. We conclude by discussing how education may affect decision-making through: personality, borrowing behavior, discount rates, risk-aversion, and the influence of employers and neighbors.