Kasl, S. V. (Dept Epidemiology and Public Health, Yale University School of Medicine, New Haven, Connecticut, U.S.A.), Cobb, S., and Gore, S. Changes in reported illness and illness behavior related to termination of employment: a preliminary report. Int. J. Epid. 1972,1:111–118. Men working in two plants which were about to shut down permanently were studied and followed-up for two years after plant closing and the resultant loss of job. Analysis of the data obtained with diverse measures of changes in self-reported health and in illness behavior (what the men did regarding such perceived changes in health) permit the following major conclusions; (1) A seasonally adjusted, general measure of complaints (days of not feeling as well as usual) revealed in both plants fluctuations over time which suggested that the index was a sensitive indicator of the stresses the men went through, such as anticipation of loss of job, periods of unemployment, and times of probationary re-employment. The men's retrospective assessments of the job loss experience and their general level of adjustment were also related to these fluctuations. (2) A seasonally adjusted general measure of disability (days of not carrying out usual activities) proved to be much less sensitive to the stresses of the situation in which the men found themselves. Additional analyses with this measure suggested that as an indicator of what a man does about this perceived state of health, it is also sensitive to aspects of the social environment (social supports) and to such personality variables as level of depression. (3) Additional measures of health and illness behavior, such as symptom checklists and frequency of use of drugs, also proved to be unsuitable as indicators of the stressful changes in the men's lives. (4) A 2-week health diary proved to be a much better approach to respondent reports of health changes than a 3-month retrospective report.