Introduction To what degree does the occupational health impact the labour market affiliation in terms of work-life expectancy? This question plays a central role in the Danish flexible labour market system with high labour force participation, relatively generous and accessible social benefits, and a low formal employment protection with a high turnover.

Method A survey of 12.429 employed Danes from 2010 is merged with longitudinal register data on social benefits to gain a follow-up period of approximately five years, but no later than the official pension age of 65 years. By using a Multi-state design and modern survival analysis, we estimate the work-life expectancy and divide the results in periods of work, unemployment, and sickness absence. The analyses are done for five occupational health scales divided into good and poor health.

Results For a 40–49 year old non-smoking women, employed in social health type of work and with elementary school education, and without the possibility of voluntary early retirement, one can expect 1.1 year decline in work-life expectancy due to poor emotional strain at work. In addition the expected time spend in sickness absence until pension age will increase by 0.9 years, and the expected time spend in unemployment will increase by 0.3 years. Similar results are made for other profiles, with or without the right to early retirement pension scheme.

Discussion Results suggest a significant influence of occupational health on work-life expectancy and work disability among Danish employees. With the new methods utilising the detailed Danish registers one have the ability to estimate work-life expectancy even for small groups, which is often the case when subgrouping levels of occupational health.