In this study results of a survey taken of persons insured under the public health care system were presented. The main objective of the survey was to determine the health of the insured on the basis of their subjective data, as well as to assess their prevention behaviour and their wish that the health insurance should offer prevention programmes. In addition, the health insurance provided data on how often the insured persons were unable to work for health reasons during a period of seven years. The data of work absenteeism are used to divide the group of respondents according to their health. To determine internal consistency it was first investigated whether persons with periods of disablement during one year (cross-section) also had long periods of disablement in the years before or after (longitudinal section). It was found that in the cross-section, persons with serious health problems (more than six weeks of disablement) differed little from those with average disablement periods (from one to six weeks) in the long-term development of disablement. The study also examined the correlation of periods of disablement and subjective data (validity). A mean correlation was found between the respondents' subjective assessment of their health and the length of disablement periods. There was also a significant negative correlation between internal health locus of control beliefs and the length of disablement. The results are discussed, and the ways and limits of using periods of disablement for empirical studies are presented.