When the goal of preventing adverse health outcomes is kept in mind, this review poses the idea that "predisease" as a category on which to act makes sense only if the following 3 conditions are met. First, the people designated as having predisease must be far more likely to develop disease than those not so designated. Second, there must be a feasible intervention that, when targeted to people with predisease, effectively reduces the likelihood of developing disease. Third, the benefits of intervening on predisease must outweigh the harms in the population.