This experiment addressed the question of whether children''s own emotional states influence their accuracy in recognizing emotional states in peers and any motives they may have to intervene in order to change their peers'' emotional states. Happiness, sadness, anger, or a neutral state were induced in preschool children, who then viewed slides of other 4-year-old children who were actually experiencing each of those states. Children''s own emotional states influenced only their perception of sadness in peers. Sad emotional states promoted systematic inaccuracies in the perception of sadness, causing children to mislabel sadness in peers as anger. Children had high base rates for using the label happy, and this significantly enhanced their accuracy in recognizing that state. Low base rates for labeling others as in a neutral state reduced accuracy in recognizing neutrality. Children were generally motivated to change sad, angry, and neutral states in peers, and they were most motivated to change a peer''s state if they were to be the agent of such change. The results are discussed in terms of the limited role of children''s own emotional states in their recognition of emotion in others or motives to intervene and in terms of factors influencing the perception of emotion, such as base rate preferences for labeling others as experiencing, or not experiencing, particular emotional states.