Relative to their young counterparts, older adults are poorer at recognizing facial expressions. A 2008 meta-analysis of 17 facial emotion recognition data sets showed that these age-related difficulties are not uniform. Rather, they are greatest for the emotions of anger, fear, and sadness, comparative with happiness and surprise, with no age-effect found for disgust. Since then, there have been many methodological advances in assessing emotion recognition. The current comprehensive meta-analysis systematically tested the influence of task characteristics (e.g., photographs vs. videos). The meta-analysis included 102 data sets that compared facial emotion recognition in older and young adult samples (N = 10,526). With task type combined, the pattern of age-effects across emotions was mostly consistent with the previous meta-analysis (i.e., largest age-effects for anger, fear, sadness; no effect for disgust). However, the magnitude and direction of age-effects were strongly influenced by elements of task design. Specifically, videos produced relatively moderate age-effects across all emotions, which indicates that older adults may not exhibit a positivity effect for facial emotion recognition. For disgust recognition, older adults demonstrated superior accuracy to young adults for the most common image set (Pictures of Facial Affect). However, they were poorer than young adults at recognizing this emotion for all other stimulus formats and image sets, which suggests that they do not retain disgust recognition. We discuss the implications that such diversity in the age-effects produced by different facial emotion recognition task designs has for understanding real-world deficits and task selection in future emotion recognition studies.