Background: Studies that differentiate among diagnoses have detected divergent results in the experience of family burden. Aim: This study aimed to investigate differences in family burden and participation in care between relatives from subgroups of psychoses, affective disorders and ‘other diagnoses’, and between different subgroups of relatives. Method: In a Swedish longitudinal study performed in 1986, 1991 and 1997, 455 close relatives of both committed and voluntarily admitted patients were interviewed concerning different aspects of their burden, need for support and participation in the actual care situation. Results: Relatives showed burdens in several of the aspects measured. In only one aspect of the investigated burden items was a difference found between different diagnostic subgroups. The relatives of patients with affective disorder more often had to give up leisure time. However, spouses showed more burdens and more often experienced sufficient participation in the patient's treatment than other subgroups while siblings more seldom experienced burdens and more seldom felt that their own needs for support had been met by the psychiatric services. Within each diagnostic subgroup there were differences between subgroups of relatives. Conclusion: Being a close relative, and living together with a severely mentally ill person in an acute situation, is one factor of importance for experiencing burden and participation in care, contradicting the conventional wisdom which differentiates between diagnoses.