Background: When studying how humans regulate their affect, it is important to recognize that affect regulation does not occur in a vacuum. As humans are an inherently social species, affect plays a crucial evolutionary role in social behavior, and social behavior likewise assumes an important role in affect and affect regulation. Emotion researchers are increasingly interested the specific ways people help to regulate and dysregulate one another’s affect, though experimental examinations of the extant models and theory are relatively few. This thesis presents a broad theoretical framework for social affect regulation between close others, considering the role of attachment theory and its developmental foundations for social affect regulation in adulthood. Affectionate and responsive touch is considered a major mechanism of regulatory benefit between people, both developmentally and in adulthood, and is the focus of the present investigation. Method: A total sample of 231 heterosexual married couples were recruited from the community. Participants were assigned to engage in affectionate touch or sit quietly, and/or engage in positive conversation prior to a stress task. Physiological data was collected continuously across the experiment. Hypotheses: Phasic respiratory sinus arrhythmia (RSA) was used to index the degree of regulatory engagement during the stressor for those who did and did not touch. It was hypothesized that touch would reduce stress appraisal and thus the need for regulatory engagement. This effect was predicted to be greater for those more anxiously attached while increasing the need for regulatory engagement in those more avoidantly attached. Secondarily, partner effects of attachment on sympathetic activation via pre-ejection period (PEP) change were tested. It was predicted that both attachment dimensions would predict a decrease in partner PEP change in the touch condition, with avoidant attachment having the strongest effect. Results: Hierarchical linear modeling techniques were used to account for nonindependence in dyadic observations. The first set of hypotheses were not supported, while the second set were partially supported. Wives’ avoidance significantly predicted husbands’ PEP change, but in the positive direction. This effect also significantly increased in the touch condition. Theoretical considerations and limitations are discussed.