The author discusses the problem of estimating the proportion of the population at risk of nutrient deficiency. While different analytical approaches to the problem have been proposed, it appears that, ideally, such an estimate would require knowledge about the joint distribution of usual intakes and nutrient requirements. The author proposes an alternative method in which it would suffice to know the mean of the requirement distribution (EAR) together with the distribution of usual intakes in the population. The paper outlines the assumptions that must hold in order for this method to be effective.