David Lewis has long defended an analysis of counterfactuals in terms of comparative similarity of possible worlds. The purpose of this paper is to reevaluate Lewis’s response to one of the oldest and most familiar objections to this proposal, the future similarity objection.