The winemaking technique of saignée is common for some varietals, and the ensuing fla- vor profiles have been carefully analyzed by oenologists. However, we argue that saignée is fundamentally about economic tradeoffs between the quantity of primary wine that is ulti- mately produced, the quality (and thus, price) of that wine, and the amount of rosé wine that is bled off in the process. We develop the first theoretically-grounded economic model of saignée and analyze the model to shed light on the winemaker’s optimal choice of saignée, and on the properties of wine and wine markets that should empirically give rise to more, or less, saignée. The model helps to explain several real-world regularities such as the absence of saignée for most Bordeaux wines, the specialization in rosé for many wines in Provence, and the practice of moderate amounts of saignée for varietals such as grenache and pinot noir.