The ability to accurately determine the emotional state of others is critical for successful social functioning. However older adults can demonstrate selective difficulties in identifying negative emotions from faces while the ability to identify positive emotional faces is preserved. In younger adults the categorization of facial expressions of emotion has been shown to rely on the processing of specific subsets of visual information (e.g. broadly smiling mouth in happiness, wide open eyes in fear)1,2. However it remains unclear whether healthy older adults process the same specific visual cues as younger adults in a less efficient manner when processing negative expressions, or if they attend to and encode qualitatively different information. We investigated whether the diagnostic information underlying the correct emotion categorization of five basic facial expressions (happy, fear, disgust, anger and sadness) changes as a function of observer age (young vs. older adults) and facial age (young vs. middle-aged vs. older faces). We applied the bubbles reverse correlation methodology with two groups of participants: younger (N=15, 18-35 years) and older adults (N=15, 65+ years). Results revealed that younger and older adults used qualitatively equivalent information to accurately categorize happy and fearful faces, but that the information used by both groups differed when they categorized fear in younger vs. older adult face stimuli. Older adults generally experienced more difficulty with the remaining negative emotions (disgust, anger and sadness) and exhibited a sub-optimal use of the diagnostic facial features. These results constitute a novel, highly detailed account of the specific visual features underlying the classification of facial expressions as across observer and transmitter of the emotional expressions stimulus age. 1Smith Cottrell, Gosselin & Schyns, 2005, Psychological Science 2Smith & Merlusca, 2014, Emotion Meeting abstract presented at VSS 2015.