The objective of this article is to discuss our research into the validation of a career-related locus of control scale. The instrument, the Vocational Locus of Control Scale developed by Fournier, Jeanrie, and Drapeau (1996), attempts to confirm the empirical existence of our conception of a five-level locus of control in the career domain. In this study, the empirical qualities of the instrument are confirmed through the demonstration of the factor analysis structure, the evaluation of inter-type correlations and the relationship established with Rotter's scale. The results, stemming from a study in which over 1,000 individuals were interviewed, demonstrate the sufficient reliability of the instrument and provide significant evidence for its conceptual validity, supporting, at the same time, the suggested nuances in the expression of a feeling of control. Questions of a theoretical and empirical nature are raised with regard to the structure of the concept.Journal of Career Assessment Volume 7/Number 1/Winter 1999/Pages 63-89The objective of this article is to discuss our research into the validation of a career-related locus of control scale. This instrument, the Vocational Locus of Control Scale (VLOC) takes its inspiration from a locus of control typology developed by Fournier, Pelletier, and Pelletier (1993) which suggested the existence of five-level continuum to express external or internal control. The typology belongs to a new generation of literature examining this concept which, on the one hand, proposes a less dichotomous view of the locus of control and which, on the other, suggests the existence of perceptions of control which vary according to the situation. The instrument developed by Fournier, Jeanrie, and Drapeau (1996) attempts to confirm the empirical existence of this subtler view of the concept, as applied to the career domain. The empirical qualities of the instrument are confirmed in this study through the demonstration of the factor analysis structure, the evaluation of inter-type correlations and the relationship established with Rotter's scale.As early as 1954, Rotter had underlined that the probability of a behaviour satisfying a need is dependent on two variables, namely the expectation of a specific reinforcement from this behaviour and the value that the Correspondence concerning this article and requests for offprints should be addressed to Professor Genevi6ve Fournier, D6partement d'orientation, d'administration et d'evaluation en education,