Abstract
Professional burnout among professionals in human services has usually been studied by making use of the Maslach Burnout Inventory (MBI). Using confirmatory factor analysis, the present study evaluates the factorial validity of the three versions of this scale that are available for Dutch-speaking regions (the full version with 22 items, a version with 20 items and a version with 16 items). The sample consisted of 995 Flemish special educators. The version of the MBI with 16 items proved to be slightly better than the others, but the fit of the model was still unsatisfactory. Implications for further research are discussed.