Until now, the instruments for the clinical evaluation of impulsivity have relied essentially on either personality inventories or to a lesser extent on specific questionnaires. Therefore this paper presents the preliminary results of a new rating scale: the Impulsivity Rating Scale (IRS). This is a 7-item hetero-evaluation of impulsivity based on the behaviour of the patient in usual life situations. The administration of the scale is easy and short: 15 minutes. The IRS was tested in five different population samples: 31 impulsive inpatients, 36 adults with major depressive episode, 15 healthy control subjects, 56 smokers before and after one week of tobacco withdrawal, and 47 adolescent depressed inpatients before and four weeks after antidepressant treatment. The results show good construct validity, good concurrent validity, good inter-rater reliability and sensitivity to change. A threshold of 8 for the total score (range 0 to 21) gives good specificity and sensitivity. The principal component analysis shows the existence of a main factor composed of all items, with lower correlation for two items which may belong to a second factor needing further investigations.