The desirability of what actually occurs is often influenced by what could have been. Preferences based on such value dependencies between actual and counterfactual outcomes generate a class of problems for orthodox decision theory, the best-known perhaps being the so-called Allais paradox. In this article we solve these problems by extending Richard Jeffrey’s decision theory to counterfactual prospects, using a multidimensional possible-world semantics for conditionals, and showing that preferences that are sensitive to counterfactual considerations can still be desirability-maximizing. We end the article by investigating the conditions necessary and sufficient for a desirability function to be a standard expected-utility function. It turns out that the additional conditions imply highly implausible epistemic principles.