Within the context of a prospective survey project among 20,000 volunteers working for "Electricité de France--Gaz de France" Company (GAZEL cohort), we performed a feasibility study using four self-assessment questionnaires: the Jenkins Activity Survey (measuring type A behaviour pattern), the Buss and Durkee Hostility Inventory, the Center for Epidemiologic Studies--Depression Scale and a personality questionnaire developed by Grossarth-Maticek and Eysenck (GME). Questionnaires were sent twice by mail, at an interval of three months, to a random sample of 408 males: 71% returned the filled forms the first time, 66% the second time, and 80% at least once. The test-retest correlation coefficients ranged from 0.62 to 0.87. The validity of the six personality type scores, calculated from GME, was tested by correlations with the scores of the three other instruments. These correlations appeared consistent with the proposed typology, especially for the "Healthy" type, the "Cancer prone" type and the "Coronary Heart Disease prone" type. The labelling of these types was justified by their association with different disease events, suggested from previous studies.