Cross-sectional findings that depressive symptoms in one older spouse influence those of the other and that marital closeness increases the influences (R.B. Tower & S.V. Kasl, 1995) were tested longitudinally. Independent interviews in 1982, 1985, and 1988 with spouse-pairs who participated in the Established Populations for the Epidemiologic Study of the Elderly showed that changes in depressive symptoms in one older spouse contributed to changes in depressive symptoms in the other. For wives in 1985 and for husbands in 1988, a spouse's baseline depressive symptoms also contributed independent variance to an increase in respondent's score on the Center for Epidemiologic Studies Depression Scale. These findings were stronger when a couple was close. These results held when known intrapersonal risk factors and the health status of the spouse were controlled.