Many daily social interactions take place digitally, for example via video calls or texts. Such digital interactions – containing limited physical cues – may yield distinct effects on interaction quality and post-interaction emotional states compared to face-to-face exchanges. In this study, we examine four preregistered hypotheses on the within-person interplay between interaction mode (i.e., digital vs. face-to-face interactions), interaction quality, and momentary well-being. We use data from two samples (N1 = 216, N2 = 21) with T1 = 5116 and T2 = 1348 Ecological Momentary Assessments (EMA) of interaction mode, interaction quality, and different momentary well-being measures (i.e., affect valence, loneliness, positive affect, negative affect). Estimating multilevel mediation Structural Equation Models, we find no within-person differences in interaction quality between digital and face-to-face interactions in Sample I. Contrary to our expectations, the quality of interaction in Sample II was rated higher during digital interactions than face-to-face interactions. Interaction quality was positively associated with momentary well-being measures in both samples. Momentary well-being was higher after face-to-face interactions than digital interactions in Sample I, but not in Sample II. Interaction quality did not mediate the relationship between interaction mode and momentary well-being – rather, it moderated the relationship in Sample I. As we do not observe evidence for a mediating effect, it remains an open empirical question through which pathways digital social interactions differ from face-to-face interactions in their effect on momentary well-being.