Objective It has been proposed that inflammation may be causally related to depression. If this is the case, it may be possible to distinguish an inflammatory depressive subtype according to illness course, pattern of co‐morbidity and symptom profile. Methods Eight hundred and eleven community dwelling older adults with depression (8 item Center for Epidemiologic Studies scale ≥ 4) from the English Longitudinal study of Ageing (ELSA) were followed for a median of 47 months. Participants with depression and inflammation (C Reactive Protein > 3 mg/l) were compared to those with depression alone. Results In a longitudinal analysis, depression with associated inflammation was more likely to persist over time. This association was independent of baseline depression severity and medical co‐morbidity (OR 1.47 95% CI 1.03 – 2.10, p = 0.034) but was no longer significant following further adjustment for body mass index (OR 1.37 95% CI 0.94 – 2.01, p = 0.106). Inflammation either partially or completely mediated the association between medical co‐morbidity, body mass index and depression at follow‐up. Depression with inflammation was associated with more amotivation, less sadness, greater medical co‐morbidity and higher body mass index. Conclusions Our findings provide some support for an inflammatory contribution to depression. This subgroup has a worse prognosis and may benefit from interventions targeting co‐morbidity, body mass index and associated inflammation. Copyright © 2016 John Wiley & Sons, Ltd.