Conceptual models of psychosocial influences on short-term changes (i.e., reactivity) in vagally-mediated heart rate variability (vmHRV) emphasize self-regulatory effort and social threat versus comfort. However, these two general perspectives have been tested separately in nearly all cases, limiting conclusions about the relative importance or possible interactive effects of effortful self-regulation and social stress. The present study compared effects of effort to regulate emotional expression and social stress versus safety on vmHRV reactivity during an interpersonal interaction, in a 2 (self-regulate emotion vs. express emotion freely) x 3 (positive vs. neutral vs. negative interaction valence) x 2 (male vs. female) between-subjects randomized factorial design. A sample of 180 undergraduates (90 women; 69 % White) discussed a current events topic (i.e., human-caused climate change) with a prerecorded partner, presented as a live interaction over a computer. Self-reports of affective responses, self-regulation effort, and appraisals of the partner\'s behavior, as well as observer ratings of participants\' behavior during the interaction, supported the effectiveness of self-regulation and interaction valence manipulations, although the former manipulation may have been somewhat weaker than the latter. Primary analyses of high-frequency heart rate variability (HF-HRV) and root mean square of successive differences (RMSSD) in heart beat intervals recorded at baseline and during the interaction revealed larger decreases in vmHRV during negative than neutral or positive interactions, but no effects of self-regulation instructions. Overall, results indicated more robust effects of social stress on vmHRV reactivity, relative to effects of self-regulatory effort.