Aims: Work disability and sickness absence increase following partner's retirement, which similarities in spouses' health could explain. We therefore studied the risk of work disability within couples, taking account of baseline health, lifestyle and socioeconomic factors.Methods: A cohort of 12,511 couples from the HUNT Study (aged 20-67 in HUNT2 1995-97) was linked to national registries, identifying all new cases of disability pension until December 2007. Data were analysed with discrete time multilevel logistic regression and Cox regression models. Partners' disability pension was included as a time-varying covariate.Follow-up time was split to examine the association dependent of time. Analyses were adjusted for age only, adjusted for health, and for lifestyle and education along with health.Results: About 15 % of an individual's propensity to receive a disability pension could be attributed couple similarity. There was an increased risk of work disability following the spouse's disability retirement [HR (hazard ratio) 1.43 (95% confidence interval 1.20-1.71) for men, HR 1.49 (95% confidence interval 1.28-1.74) for women]. The association was somewhat attenuated after adjustments for health, lifestyle and education.Conclusion: There was a substantial clustering of disability pensions within couples, which cannot be explained by similarities in health, lifestyle and education. This suggests partner's influence each other's work ability. From a clinical perspective, the family situation needs to be taken into account when addressing health promotion and work participation.