Psychosocial factors predict the development and course of cardiovascular disease, perhaps through sympathetic and parasympathetic mechanisms. At rest, heart rate (HR) is under parasympathetic control, often measured as high-frequency heart rate variability (HF-HRV). During stress, HR is influenced jointly by parasympathetic and sympathetic processes, the latter often quantified as pre-ejection period (PEP). In studies of cardiovascular risk factors that involve social interaction (e.g. marital conflict), HF-HRV might be altered by speech artifacts, weakening its validity as a measure of parasympathetic activity. To evaluate this possibility, we tested associations of HF-HRV and PEP with HR at rest and across periods of marital conflict interaction that varied in experimentally-manipulated degrees of speech in 104 couples. At rest, only HF-HRV was independently related to HR, for both husbands and wives. During speaking, listening, and recovery periods, husbands\' and wives\' HF-HRV and PEP change independently predicted HR change. These findings support interpretation of HF-HRV as a parasympathetic index during stressful social interactions that may confer risk for cardiovascular disease.