Previous studies have found a relationship between attributions of blame and traumatic events such as crime, illness, and accidents/disasters, albeit inconclusive as to the benefits or detriments of self- and other-blame on adjustment outcome (e.g., Janoff-Bulman, 1979; Joseph, Brewin, Yule & Williams, 1991,1993; Derry & McLachlan, 1995; Frazier & Schauben, 1994). The effects of attributions of blame on the adjustment outcome of family members bereaved through murder has been neglected. Therefore, little is known about such benefits to adjustment in this population. In addition, no longitudinal research has been conducted so little is known about this process of adjustment. A retrospective longitudinal study investigated emotional state and event-related impact, attributions of blame, control and just world cognitions, revenge and disabling distress. Thirty-four family members, recruited from "Families of Murdered Children", were interviewed and completed four psychological measures. They were followed up six and twelve months later. On all three occasions, subjects showed high levels of negative emotional state and event-related impact, especially older, female and support seeking subjects. Self-blame and feelings of revenge were linked to higher levels of negative emotional state and event-related impact, especially in female subjects. Control and just world cognitions were not related to emotional state and event-related impact. Negative emotional state at Time 1 was predictive of poor overall adjustment at Time 2 and Time 3, while gender was predictive of poor overall adjustment at Time 2. Subjects suffering from distress that interfered with their daily lives at Time 3 had higher negative emotional state and event-related impact at Time 1, Time 2 and Time 3. In order to further investigate the effects of blame attributions on mood, a randomised between-subjects laboratory study was conducted. Eighty-seven undergraduates were assigned to one of three writing conditions (self-blame, other-blame and no blame/control) with mood being assessed before and after writing. Results showed that negative mood had been cognitively induced, however, no condition effects occurred. The mood effect was greater for women than men. Implications for theory, practice and future research in relation to the main findings are discussed.