BackgroundManagement of long-term depression is a significant problem in primary care populations with considerable on-going morbidity, but few studies have focused on this group.AimsTo evaluate whether structured, nurse-led proactive care of patients with chronic depression in primary care improves outcomes.MethodParticipants with chronic/recurrent major depression or dysthymia were recruited from 42 UK general practices and randomised to general practitioner (GP) treatment as usual or nurse intervention over 2 years (the ProCEED trial, trial registration: ISRCTN36610074).ResultsIn total 282 people received the intervention and there were 276 controls. At 24 months there was no significant improvement in Beck Depression Inventory (BDI-II) score or quality of life (Euroquol-EQ-VAS), but a significant improvement in functional impairment (Work and Social Activity Schedule, WSAS) of 2.5 (95% CI 0.6–4.3, P = 0.010) in the intervention group. The impact per practice-nurse intervention session was –0.37 (95% CI –0.68 to –0.07, P = 0.017) on the BDI-II score and –0.33 (95% CI –0.55 to –0.10, P = 0.004) on the WSAS score, indicating that attending all 10 intervention sessions could lead to a BDI-II score reduction of 3.7 points compared with controls.ConclusionsThe intervention improved functioning in these patients, the majority of whom had complex long-term difficulties, but only had a significant impact on depressive symptoms in those engaging with the full intervention.