In Europe, several studies are currently underway to investigate the cancer risk of pilots and cabin crew exposed to low-level ionizing radiation of cosmic origin. Although no individualized exposure measurements of airline personnel are available, exposure assessment based on job history data is feasible. However, there is a marked variability in the level of detail of these data between studies in different countries and between subcohorts in national studies raising the issue of comparability of exposure estimation. In this paper we investigate the comparability of several methods of exposure assessment in a large German cohort of pilots and cabin crew. We found that the correlation between the estimates obtained by the four approaches analysed, is relatively high, ranging from 0.85 to 0.97. The precision attainable in the exposure assessment is higher than in many other epidemiological studies but can be refined further with simulation studies and comparison with ongoing and future on-board measurement programmes.