Abstract Epidemiological differences in rates of mood disorders – higher prevalence among females and those of lower socioeconomic status – can be traced to intermediate psychosocial processes such as stressful life events. Greater vulnerability to such stressors has been traced to lack of support in responding to these, both currently in adulthood and in childhood, the latter particularly associated with low self-esteem in adulthood. Latterly such vulnerability to stressors has also been linked with a functional polymorphism in the promoter region of the serotonin transporter gene, with recent evidence of this gene–environment interaction. Preventive psychosocial interventions providing social support have been developed based upon this aetiological model. One successful randomized controlled trial involving one-to-one volunteer befriending for chronic depression has recently been repeated in another RCT prospective on perinatal depression. Mothers’ groups based around visits to GP surgeries are described as possible extensions of such services.