In this proof-of-principle study, a convenience sample of 55 dyadic psychotherapy sessions conducted by one therapist was analyzed. The electrocardiograms and respiration behavior of both therapist and client were monitored simultaneously. Four clients were included, and session outcome was documented by session reports in two clients. From electrocardiograms, heart rate and heart rate variability were derived in consecutive 15-second intervals throughout sessions. Entire sessions (average duration, 51 minutes) were assessed for physiological synchrony of therapist's and client's respiration, electrocardiogram, heart rate, and heart rate variability. Two methods of synchrony computation were applied to the time series: windowed cross-correlation and correlation of local slopes (concordance). Both methods included surrogate controls using segment-wise shuffling. Significant synchrony of three measures, but not of electrocardiograms, was present in this dataset. In regression models, we found associations between synchronies and alliance ratings, and further selfreport variables. Results support the existence of physiological synchrony in this collection of psychotherapy sessions, which speaks for the sympathetic and parasympathetic coupling between this therapist and her client and its link with ratings of therapy process. The feasibility of deriving signatures of synchrony of physiological signals with the described methodology was corroborated. The findings now await generalization by further research.