Romantic relationships are important contexts for substance use and emotional well-being. We tested the hypotheses that (i) genetic predispositions for alcohol consumption would be positively associated with partner substance use, (ii) partner substance use would moderate genetic influences on one’s own alcohol outcomes, and (iii) partner discordance in substance use would be associated with lower emotional well-being and relationship quality.

Analyses included 2,357 participants (Mage = 51.4, 58.2% female) from the Collaborative Studies on the Genetics of Alcoholism. Focal measures included participants’ reports of their own and their current partner’s past-year substance use (frequencies of alcohol use, heavy drinking, drunkenness, cannabis use, and nicotine use), emotional well-being, and relationship quality. Participants’ genetic predispositions were indexed with genome-wide polygenic scores for alcohol consumption (PGSAlc). Participant–partner substance use discordance was calculated as the difference between the participant’s and their partner’s use for each substance use measure, separately.

Participant PGSAlc was not significantly associated with partners’ perceived substance use. Frequent perceived partner alcohol use and heavy drinking significantly amplified the association between PGSAlc and alcohol use or drunkenness. Frequent perceived partner drunkenness and cannabis use significantly attenuated the association between PGSAlc and heavy drinking or frequency of alcohol use. Participant–partner discordance for several substance use measures was significantly associated with lower emotional well-being and relationship quality, controlling for participant and partner substance use main effects.

The results highlight the importance of partner substance use in etiological models of alcohol use, emotional health outcomes, and relationship quality.