In this paper an entropy model with variable target is presented, in which target values are assumed to belong to a specified convex set, so that multiple base-year information and forecasts of future trends can be handled without prior aggregation of such information into one fixed target. Three solution methods for such a model are presented—one cutting-plane and two search methods—all of which utilize the fact that entropy models with fixed targets can be solved efficiently. Some comments on potential applications and reasonable target-set specifications are made and numerical examples are presented.