One potentially important drawback of existing theories of limited attention is that they typically assume a rich dataset of choices from many menus. We study the problem of identifying the distribution of cognitive characteristics in a population of agents when only aggregate choice behavior from a single menu is observable. We show how both “consideration probability” and “consideration capacity” distributions can be substantially identified by aggregate choice shares. We also suggest how to embed the attention models in an econometric specification of the inference problem. Finally, we successfully use our results to recover the true parameters in Monte Carlo simulations of both models.