We examined whether self-biases in perceptual matching reflect the positive valence of self-related stimuli. Participants associated geometric shapes with either personal labels (e.g., you, friend, stranger) or faces with different emotional expressions (e.g., happy, neutral, sad). They then judged whether shape–label or shape–face pairs were as originally shown or re-paired. Match times were faster to self-associated stimuli and to stimuli associated with the most positive valence. In addition, both the self-bias and the positive emotion bias were reliable across individuals in different test sessions. In contrast there was no sign of a correlation between the self-bias and the emotion-bias effects. We argue that self-bias and the bias to stimuli linked to positive emotion are separate and may reflect different underlying processes.