Independent interviews with husbands and wives from 317 community-dwelling older couples showed that depressive symptoms in 1 spouse had a significant impact on depressive symptoms in the other spouse, after controlling for potentially confounding respondent sociodemographic and health status variables. Relationship quality moderated the influence, with spouse depressive symptoms contributing more of the variance to a respondent's symptoms when a couple were close than when they were not and closeness to a spouse buffering the potentially depressive effects of a respondent's own frailty and financial distress. Husband's own variables explained more of the variance in his symptoms when the couple were not close than when they were, and closeness to his wife increased a husband's vulnerability to the impact of the wife's health status.