The ‘behavioral pattern A’ and the ‘B’ pattern in selected groups of male populations in the framework of a prospective and predictive study of a large population of civil service employees is compared. The Personality Inventory of Lazare, Klerman and Armor and the auto-questionnaire of Bortner were used to differentiate the ‘A’ and ‘B’ group populations. Selected cards from Murray’s TAT seem more appropriate in this respect and demonstrate the lack of efficiency of certain questionnaires in areas of psychosomatic research. Moreover the selected TAT card responses provide some evidence of ‘pensée opératoire’ or alexithymia in ‘A’ group subjects whatever its metapsychological implication.