Purpose: The aims of this study are: to present a training evaluation based on the "four levels model"; to demonstrate the value of experimental designs in evaluation studies; and to take a first step in the development of an evidence-based training program. Design/methodology/approach: The Kirkpatrick four levels model was used to evaluate the effectiveness of management training. Hypotheses about the effects of internal/external "locus" of control and experiential learning styles were tested. Measures gathered at several points of time were compared to measures taken from a control group. "Level four" results were obtained by 360-degree feedback appraisal. Findings: Trainees' knowledge improved significantly. Trained managers also demonstrated more "good" management behaviors and they scored higher on long-term results. Trained supervisors with an internal "locus" of control acquired more knowledge than trained supervisors with an external "locus" of control. However, internals did not differ significantly from externals on actual behavior measures. The hypothesized relationship between learning style and training effectiveness was not supported. The four levels method gives detailed insights in results. Attention should be paid to trainees' "locus" of control. Originality/value: The paper outlines the value of the "four levels model" of training evaluations and contributes to the "theory-based, evidence-based" approach to the development of training programs. (Contains 6 tables and 2 figures.)