Companionship is related to better affect and relationship satisfaction, but few studies have examined both partners' perspectives over time and the link between companionship and health. In three intensive longitudinal studies (Study 1: 57 community couples; Study 2: 99 smoker–nonsmoker couples; Study 3: 83 dual-smoker couples), both partners reported daily companionship, affect, relationship satisfaction, and a health behavior (smoking in Studies 2 and 3). We proposed a dyadic score model that focuses on the couple level for companionship as a dyadic predictor with considerable shared variance. On days with higher companionship, couples reported better affect and relationship satisfaction. When partners differed in companionship, they also differed in affect and relationship satisfaction. For smoking, a different picture emerged: Whereas smokers with nonsmoking partners smoked less on average with higher companionship, smokers with smoking partners smoked more on days with higher companionship. Findings show companionship as a consequential relationship construct deserving further study. Using the dyadic score model acknowledged both partners' perspectives on companionship. It demonstrated higher precision for detecting effects of partner averages in a dyadic predictor compared with traditional approaches, tests for effects of partner differences in a dyadic predictor and in outcome while maintaining the focus on the dyad.
