BACKGROUND: The long-term impact of alcohol intake in midlife on developing depression is not clear. We aimed to investigate drinking pattern during midlife as a risk factor for developing depression during 28 years of follow-up. METHODS: We used data from a well characterised prospective cohort study (the Whitehall II study) of 7478 men and women (70% male) aged 35 to 55 years, and free from depression in 1985-1988, followed up regularly until 2013. Drinking pattern was defined in terms of usual and maximum amounts consumed within a single drinking session, total weekly volume of alcohol consumed and drinking frequency. Depression was assessed using the General Health Questionnaire Depression Subscale at multiple follow-up occasions (up to eight times in total). Associations between different drinking pattern components during midlife and depression were estimated using flexible parametric survival models. RESULTS: After adjustment for confounding factors only abstaining from alcohol during midlife was associated with an increased risk of developing depression during long-term observation. However, this association became non-significant after adjusting for baseline self-reported health. CONCLUSIONS: In this community dwelling population, drinking pattern during midlife was not associated with an increased risk of developing depression.