The ability to recognize how another is feeling is a critical skill, with profound implications for social adaptation. Training programs designed to improve social functioning typically attempt to direct attention toward or away from certain facial configurations, or to improve discrimination between emotions by categorizing faces. However, emotion recognition involves processes in addition to attentional orienting or categorical labeling. The intensity with which someone is experiencing an emotion is also influential; knowing whether someone is annoyed or enraged will guide an observer's response. Here, we systematically examined a novel paradigm designed to improve ratings of facial information communicating emotion intensity in a sample of 492 participants across a series of 8 studies. In Study 1, participants improved precision in recognizing the intensity of facial cues through personalized corrective feedback. These initial findings were replicated in a randomized-control trial comparing training with feedback to viewing and rating faces without feedback. Studies 2 and 3 revealed that these effects generalize to identities and facial configurations not included in the training. Study 4 indicated that the effects were sustained beyond the training session. These findings suggest that individualized, corrective feedback is effective for reducing error in rating the intensity of facial cues. (PsycInfo Database Record (c) 2022 APA, all rights reserved).