Most studies of meditation have focused on "actual affect" (how people actually feel). We predict that meditation may even more significantly alter "ideal affect" (how people ideally want to feel). As predicted, meditators ideally wanted to feel calm more and excited less than nonmeditators, but the groups did not differ in their actual experience of calm or excited states (Study 1). We ruled out self-selection and nonspecific effects by randomly assigning participants to meditation classes, an improvisational theater class, or a no class control (Study 2). After eight weeks, meditators valued calm more but did not differ in their actual experience of calm compared with the other groups. There were no differences in ideal or actual excitement, suggesting that meditation selectively increases the value placed on calm. These findings were not due to expectancy effects (Study 3). We discuss the implications of these findings for understanding how meditation alters affective life.