Empowerment is a widely used construct in research on social work, mental health and community interventions, but has only been exploited indirectly with the unemployed. But job finding is an important dimension of empowerment and could be used to test the accuracy of the concept and of its measures. The Making Decisions Empowerment Scale was used with 97 unemployed people who had been jobless for 6 months. Even though the psychometric qualities of the 5 subscales and the total scale were mixed, convergent and discriminant validity with several adaptive and non-adaptive dimensions could be established for the global scale and for the Esteem, Power, Control and, to a lesser degree, the Activism subscales. The results were only marginally better for the 28 items global scale compared to the 9 items Esteem scale. Empowerment could be adequately modelled by using three dimensions: change coping, depression, and chance control of unemployment. Comparing 6 months later those who had found a job with the still unemployed, the 2 groups differed significantly on 2 of the 5 subscales (Activism and Control) though not on the total empowerment scale, nor on the other psychometric scales. The results throw some doubt on the accuracy of an aggregate measure that sums up divergent dimensions. Instead, it is proposed that more specific and individualised constructs be used, at least in unemployment research.