This paper investigates the interpretation of conditionals headed by the modal would. It presents a unified account of such conditionals based on a Lewis-Stalnaker approach, and derives differences from the interpretation of aspectual heads. Perfective aspect is characterized as making reference to events, and 'anchoring' antecedent clauses to the actual world, whereas perfect aspect is characterized as quantifying over events, allowing for 'truly counterfactual' interpretations. A Lewis-style view of events is adopted to deal with the interpretation of event pronouns and to explain the matching of events across worlds. Differences in the interpretation of aspectual heads provide an explanation for differences in the intepretation of conditionals headed by would by imposing different constraints on the quantificational domain of the modal.