There are differences of opinion about self-esteem enhancement in the classroom; these differences exist at both conceptual and practical levels. The aim of this study was to ascertain whether techniques employed by primary school teachers as a day-to-day part of their teaching can have measurable effects on the self-esteem of their pupils. Two different approaches to self-esteem enhancement in primary classes were evaluated. The participants were 519 primary school children and their teachers (n = 21). Circle-Time and efficacy-based approaches were compared with a control condition over a four-month period. On two self-report measures of self-esteem, gains were found for both experimental conditions, but not for the controls. Significant sub-scale differences suggested that the two approaches achieved their effects in different ways, consistent with a two-dimensional model of self-esteem. Circle-Time methodology, focusing on the creation of a climate in which individuals are respected and valued, is more likely to help children to develop a sense of self-worth. On the other hand, an efficacy-based approach tends to focus on the achievement of performance goals, and is thus more likely to develop the self-competence dimension. It is argued that we should now reconsider how we think about self-esteem enhancement in primary classrooms. (Contains 4 tables.)