This is an experimental study of pain communication in couples. Despite evidence that chronic pain in one partner impacts both members of the dyad, dyadic influences on pain communication have not been sufficiently examined and are typically studied based on retrospective reports. Our goal was to directly study contextual influences (ie, presence of chronic pain, gender, relationship quality, and pain catastrophizing) on self-reported and nonverbal (ie, facial expressions) pain responses. Couples with (n = 66) and without (n = 65) an individual with chronic pain (ICP) completed relationship and pain catastrophizing questionnaires. Subsequently, one partner underwent a pain task (pain target, PT), while the other partner observed (pain observer, PO). In couples with an ICP, the ICP was assigned to be the PT. Pain intensity and PO perceived pain intensity ratings were recorded at multiple intervals. Facial expressions were video recorded throughout the pain task. Pain-related facial expression was quantified using the Facial Action Coding System. The most consistent predictor of either partner's pain-related facial expression was the pain-related facial expression of the other partner. Pain targets provided higher pain ratings than POs and female PTs reported and showed more pain, regardless of chronic pain status. Gender and the interaction between gender and relationship satisfaction were predictors of pain-related facial expression among PTs, but not POs. None of the examined variables predicted self-reported pain. Results suggest that contextual variables influence pain communication in couples, with distinct influences for PTs and POs. Moreover, self-report and nonverbal responses are not displayed in a parallel manner.