Self-reported work disability is analyzed in the US, the UK and the Netherlands. Different wordings of the questions lead to different work disability rates. But even if identical questions are asked, cross-country differences remain substantial. Respondent evaluations of work limitations of hypothetical persons described in vignettes are used to identify the extent to which differences in self-reports between countries or socio-economic groups are due to systematic variation in the response scales. Results suggest that more than half of the difference between the rates of self-reported work disability in the US and the Netherlands can be explained by response scale differences. A similar methodology is used to analyze the reporting bias that arises if respondents justify being on disability benefits by overstating their work limiting disabilities.