Vulnerable travellers face challenges in the transport environment potentially leading to decreased mobility. A research goal is to find out how to improve railway accessibility by reducing barriers for persons with functional limitations. A method for measuring accessibility has earlier been developed where travel barriers are assigned different weights based on persons’ functional ability and travel behaviour. The more weight placed on a certain barrier, the less accessible and thus less probable a particular journey will be. In the present work, these travel-barrier weights are analysed psychometrically based on the replies to questions about ease or difficulty of accessing travel with various barriers in responses given by 162 older long-distance travellers. An invariant measure theory approach (Rasch) is employed that allows (i) transformation of ordinal questionnaire data onto a quantitative interval scale; and (ii) separate measures of barrier level of challenge and person capability. A principal component analysis revealed three main clusters: 1) mainly ergonomically related questions; 2) mainly informational/cognitive; while cluster 3) is a mix of these two. Correlations are investigated between perceived accessibility and functional ability, and between person capability and functional ability. Independent sources of measurement uncertainty (e.g. under-estimation of scores) are distinguished from separate estimates of task challenge and individual travel capability. These independent sources are accounted for in estimates of reliability and validity of the various measures.