There is an ongoing debate concerning the contribution of different aspects of empathy to achieving an accurate understanding of others. In this study, we aimed to better comprehend the roles of experience sharing and mentalizing using a modified empathic-accuracy task. We analyzed the unique contribution of each of these mechanisms with an explicit cognitive report as well as an affective physiological synchrony measurement. First, we recorded the emotional autobiographical stories told by participants ("targets", N = 28). Then, the targets watched their own videos as their heart rate (HR) was measured, and they reported on both a continuous and a discrete emotion scale what they felt while relaying the story. Next, we collected HR data from new participants ("observers", N = 72) as they similarly rated the targets' valence and discrete emotional states. In order to test the contribution of sensorimotor cues and contextual cues to empathic accuracy, observers viewed some videos with audio, others without audio, and listened to a third set of only the audio. We hypothesized that empathic accuracy-a cognitive measure that is a proxy for mentalizing and is operationalized by the correlation between a target's reported emotions and an observer's inference of those emotions-would be greater when linguistic information is present. We also hypothesized that physiological synchrony, a proxy for experience sharing, would be greater in the video-only condition, which was limited to sensorimotor cues to infer the other's emotional state. Indeed, we found that empathic accuracy was greater when auditory information was present, and that HR synchrony was more prevalent when visual cues were presented alone. Having both information streams together did not enhance accuracy, yet it was the only condition in which both behavioral empathic-accuracy measures correlated with HR synchrony. This study provides evidence that separate experience sharing and mentalizing pathways are active in the same task.