||AIMS: To examine associations of abstention, alcohol consumption and problem drinking with subsequent disability pensioning (DP), and whether previous excessive consumption ('sick-quitting') could explain some of the increased risk for DP among abstainers. DESIGN: Prospective population-based study. SETTING AND PARTICIPANTS: Data were from two waves of the Nord-Trondelag Health Study (HUNT) linked with the national insurance database. The two main analyses included 37,729 (alcohol consumption) and 34,666 (problem drinking) participants. MEASUREMENTS: Alcohol consumption was measured by self-reported consumption, while problem drinking was assessed by the Cut down, Annoyed, Guilt, Eye-opener (CAGE) questionnaire. Information on subsequent DP, including diagnosis for which the DP was awarded, was gathered from the national insurance database. Covariates included somatic illness and symptoms, mental health, health-related behaviour, socio-economic status and social activity. FINDINGS: Those reporting the highest level of alcohol consumption were not at increased risk for DP [hazard ratio (HR) 1.12, 95% confidence interval (CI): 0.92-1.38], whereas problem drinking was a strong predictor (HR 2.79, 95% CI: 2.08-3.75) compared to their corresponding reference groups. Alcohol abstainers were also at increased risk for DP, but among them, the previous consumers (HR 1.95, 95% CI: 1.48-2.57) and previous excessive consumers (HR 1.67, 95% CI: 1.01-2.74) were at higher risk for DP than constant abstainers. CONCLUSIONS: Problem drinking is linked to subsequent requirement for a disability pension but mere alcohol consumption is not. This is partly explained by 'sick-quitting'.