BACKGROUND: Several measures have been proposed to evaluate temperament traits and their connection to psychopathology. One recent development in this area is the TEMPS-A (Temperament Evaluation of Memphis, Pisa, Paris and San Diego Auto-questionnaire). The purpose of this study is to psychometrically validate the Lebanese-Arabic version of TEMPS-A. METHODS: The TEMPS-A was adapted to Lebanese-Arabic and administered to 1320 Lebanese (593 males, 727 females, mean age of 43+/-16 years) representing the adult population of Lebanon (total population: 4.2 million). This section was added by the Institute for Development Research and Applied Care (IDRAC) to the cross-national World Mental Health (WMH) initiative Lebanese chapter. Chronbach-alpha and Pearson's correlation were used to test the internal consistency and correlation among the subscales. Factor loadings were calculated using the principal component analysis with varimax rotation. RESULTS: Internal consistency varied between 0.66 (depressive) and 0.88 (anxious). As formulated by the originators of the instrument, five main factors emerged from the factor analysis, with some overlap between the depressive, the cyclothymic and the anxious temperaments, but almost none for the hyperthymic and irritable temperaments. The strongest correlation was observed between the anxious and the cyclothymic temperament subscales. Women had higher mean scores on the depressive, cyclothymic, and anxious subscales, whereas men scored higher on the hyperthymic subscale. There was a trend of increase by age in scores of the depressive subscale, and a trend of decrease in the cyclothymic and irritable subscales. CONCLUSION: The Lebanese-Arabic TEMPS-A has a good internal consistency, is easy to use, and opens new and interesting research avenues in large epidemiologic national studies on temperament.