Physiological linkage refers to the degree to which peoples' physiological responses change in coordinated ways. Here, we examine whether and how physiological linkage relates to incidents of shared emotion, distinguished by valence. Past research has used an "overall average" approach and characterized how physiological linkage over relatively long time periods (e.g., 10-15 min) reflects psychological and social processes (e.g., marital satisfaction, empathy). Here, we used a "momentary" approach and characterized whether physiological linkage over relatively short time periods (i.e., 15 s) reflects shared positive emotion, shared negative emotion, or both, and whether linkage during shared emotions relates to relational functioning. Married couples (156 dyads) had a 15-min conflict conversation in the laboratory. Using behavioral coding, each second of conversation was classified into 1 of 4 emotion categories: shared positive emotion, shared negative emotion, shared neutral emotion, or unshared emotion. Using a composite of 3 peripheral physiological measures (i.e., heart rate, skin conductance, finger pulse amplitude), we computed momentary in-phase and antiphase linkage to represent coordinated changes in the same or opposite direction, respectively. We found that shared positive emotion was associated with higher in-phase and lower antiphase linkage, relative to the other 3 emotion categories. Greater in-phase physiological linkage during shared positive emotion was also consistently associated with higher-quality interactions and relationships, both concurrently and longitudinally (i.e., 5 to 6 years later). These findings advance our understanding of the nature of physiological linkage, the emotional conditions under which it occurs, and its possible associations with relational functioning. (PsycInfo Database Record (c) 2021 APA, all rights reserved).