Background. Studies on the direct and buffering effects of social support have not examined psychiatric sickness absence and few studies have considered support both at home and at work. This study addresses prospectively the effects of chronic stressors and social supports, at home and at work, on psychiatric sickness absence rates.

Methods. Sociodemographic factors, health and social support were measured at baseline, and short and long spells of sickness absence were measured prospectively over a 5-year period. The participants were a subsample of 4202 male and female civil servants, aged 35–55 years at baseline, from an occupational cohort, the Whitehall II Study, who completed detailed social support questions.

Results. Support from colleagues and supervisors at work is related to lower risk of short spells of psychiatric sickness absence, particularly for those also receiving high levels of negative aspects of close relationships from their closest person outside work. Negative aspects of close relationships from the closest person increase the risk of taking long spells of psychiatric sickness absence in men. High levels of material problems increase the risk of short spells of sickness absence.

Conclusions. Negative aspects of close relationships may have an aetiological role in non-psychotic psychiatric disorder. Social support at work appears to protect against short spells of psychiatric sickness absence. This potentially implies that levels of short spells of absence might be reduced by increasing support at work. Conversely, emotional support at home may influence absence-related behaviour and encourage a person to take absence at a time of illness.