We examined if persistent depressive symptoms are associated with markers of inflammation (C-Reactive Protein-CRP) and coagulation (fibrinogen), and if this association can be partly explained by weight control and behavioural risk factors (smoking, alcohol, physical activity). The study sample included 3609 men and women (aged 60.5+/-9.2 years) from The English Longitudinal Study of Ageing, a prospective study of community dwelling older adults. Depressive symptoms (using the 8-item CES-D scale), health behaviours (smoking, alcohol, physical activity), body weight, and central adiposity were assessed at baseline and 2 years follow up. CRP and fibrinogen were assessed at follow up. At baseline 12.7% of the sample reported elevated depressive symptomatology, which persisted in 6.1% of participants at follow up. Baseline CES-D score was associated with CRP (beta=.035, SE=.0066) and fibrinogen (beta=.023, SE=.0060) measured 2 years later. Using simple mediation analysis we observed both a direct association of depressive symptoms on CRP (beta=.013, SE=.0066) and indirect mediating effects through behavioural risk factors (beta for total indirect effect beta=.022, SE=.0023). For fibrinogen there were no direct effects of depression, and the association was entirely explained through indirect mediating effects of health behaviours. The presence of recurrent elevated depressive symptomatology at both time points was more strongly associated with CRP and fibrinogen. In summary, the association between depressive symptoms and low grade inflammation can be partly explained by behavioural risk factors. The presence of persistent depression appears to be associated with the greatest risk of elevated inflammation.