The current longitudinal study examined the consequences of spouses' tendencies to forgive their partners over the first 2 years of 72 new marriages. Though positive main effects between forgiveness and marital outcomes emerged cross-sectionally, spouses' tendencies to forgive their partners interacted with the frequency of those partners' negative verbal behaviors to predict changes in marital outcomes longitudinally. Specifically, whereas spouses married to partners who rarely behaved negatively tended to remain more satisfied over time to the extent that they were more forgiving, spouses married to partners who frequently behaved negatively tended to experience steeper declines in satisfaction to the extent that they were more forgiving. Similar patterns emerged for changes in the severity of husbands' problems, such that husbands married to wives who frequently behaved negatively reported sharper increases in problem severity to the extent that they were more forgiving but reported more stable problem severity to the extent that they were less forgiving. These findings question whether all spouses should benefit from forgiveness interventions and thus highlight the need for further research on the most appropriate targets for such interventions.