Alcohol use disorders (AUD) are often comorbid with depressive symptoms. Cohort studies on the association between AUD and subsequent depressive symptoms have produced inconsistent results. Moreover, regarding alcohol intake, the risk of developing depressive symptoms might vary with alcohol intake level. We aimed to investigate the association between AUD, alcohol intake and subsequent depressive symptoms. We conducted a systematic search in PubMed, Embase and PsycINFO for cohort studies on the association between AUD or alcohol intake and subsequent depressive symptoms. We included 338 426 participants from 42 studies. Six and four studies analyzed only females and males, respectively. We combined risk estimates for developing depressive symptoms using a random-effects model. We divided alcohol intake into abstinence, light (0-84g/week), moderate (85-168g/week) and heavy drinking (>168g/week or>48g/day at least weekly). We conducted a categorical analysis to compare the risk of depressive symptoms between abstinence and different intake categories. Further, we conducted a dose-response analysis to investigate the alcohol-depression association. We analyzed 42 studies (follow-up time: 1-40years). AUD was associated with significantly increased risk of subsequent depressive symptoms [relative risk (RR)=1.57, 95% confidence interval (CI)=1.41-1.76]. Regarding alcohol intake, heavy drinking had an increased risk of depressive symptoms; however, the association was only significant when controls were limited to non-heavy drinkers (RR=1.13, 95% CI=1.05-1.22). Taking into consideration the possibility of publication bias and confounding factors made the association non-significant. We observed J-shaped associations in both categorical and dose-response analyses where light-moderate drinking had a significantly decreased risk of depression, while heavy drinking did not show a significant association with depressive symptoms compared with non-drinkers. Alcohol use disorders are associated with increased the risk of subsequent depressive symptoms. Heavy drinking does not significantly predict occurrence of depressive symptoms after adjusting for potential confounders.