The neuropeptide oxytocin (OT) may be beneficial to augmenting behavioral therapies for couples given its ability to alleviate stress reactivity and increase pro-social behavior. However, there is growing evidence demonstrating inconsistent effects of OT on social behavior. Infidelity may represent a relational vulnerability that modulates the OT response. The present study hypothesized that actor and partner emotional and physical infidelity would be associated with less adaptive conflict behaviors, and moderate the OT response, such that participants randomized to a OT condition, compared to placebo, who report more infidelity would show less adaptive conflict behaviors. Participants were 30 couples ( = 60) wherein one or both partners engaged in recent hazardous drinking or illicit drug use. Partners completed a 10-min dyadic conflict task in the laboratory, self-administered a single intranasal dose of OT (40 IU) or placebo, and then completed the same 10-min dyadic conflict task following a 45-min wait period. Couple conflict behaviors were observed during the conflict tasks and assessed using a validated coding system. Actor partner interdependence models detected significant interactions between drug condition and physical infidelity, such that among individuals in the OT group, verses individuals in the placebo group, (a) who reported greater physical infidelity had greater distress maintaining attributions and (b) whose partners reported greater physical infidelity had fewer relationship enhancing attributions. Results are consistent with the social salience hypothesis of OT and suggest that physical infidelity appears to serve as a contextual vulnerability that may decreases positive and increase negative behaviors during conflict. (PsycINFO Database Record (c) 2019 APA, all rights reserved).