Political leaders use emotions to attract support from voters, take positions on issues or policies, and enhance communication. Yet men and women are constrained by gender role expectations, which limit the range and type of acceptable emotions that a leader can express. In this paper, we examine how gender shapes the behavior of political leaders through a new venue: emotional expression in images posted by leaders on social media. We use more than a million posted images from all U.S. Congressional social media accounts (including Twitter, Instagram, and Facebook). For all images posted, we use supervised machine learning methods to identify anger, sadness, and happiness in the faces in images posted by Members of Congress (MOCs). Specifically, we 1) develop a custom facial recognition model to identify MOCs in social media imagery and 2) train and validate a convolutional neural network (CNN) to detect emotional displays. Drawing on research on role congruity, we argue that women MOCs will post images with a broader range and higher level of emotions, but will be less likely than men MOCs to post images of themselves expressing emotion. The project offers an opportunity to understand how political leaders use emotions and images to communicate to constituents and gender shapes these strategic actions.