Premature discontinuation presents challenges to couple clients and therapists alike. Although couple therapy has demonstrated efficacy and effectiveness, little is known about the mechanisms that contribute to couple therapy completion. This study presents the results of an observational inquiry into the psychotherapeutic processes associated with treatment discontinuance using clinical data. Using observational coding of the Rapid Marital Interaction Coding System (RMICS) researchers sought to examine differences in in‐session interactions between couple members in a matched sample of therapy continuers and discontinuers. Results indicate that specifically for female partners, the absence of positive interactions is linked to therapy discontinuation. Clinical implications of these findings are presented and discussed.