The evaluation of a management training programme for research scientists and senior technicians formed the basis of this applied research project. Before, during and after measures were used to assess the knowledge, skills, attitudes and work behaviours of trainees and control group members in order to compare the changes which occurred as a result of the programme. Both formative and summative evaluation were incorporated into the experimental design and careful attention was paid to internal and external validity issues by the use of matched control groups, multiple measures of the dependent and independent variables, including both objective and subjective methods, self-reports and reports by trainees' superiors. Repeated measures of knowledge, skills, attitudes and work behaviours were taken utilizing a longitudinal design and some replication was conducted with other groups of trainees at other times. The plan of approach involved an initial analysis of training needs, examination of the input to training including course content and training techniques, immediate and longer term evaluation of subjects' reactions to the programme and an assessment of the outcomes of training. Outcome evaluation comprised measures of effectiveness at both immediate and intermediate levels with measurements taken immediately and at three, six and twelve months following training. Effectiveness was assessed in relation to the goals determined during the preliminary analysis of training needs, the focus shifting gradually during the course of evaluation from an emphasis on learning effects to changes in work behaviour and performance. Formative evaluation in the form of feedback to course controllers was provided throughout while the summative evaluation consisted of a final reporting of the effectiveness of the training programme over a twelve month period and the consistency of results from a subsequent sample of trainees. A model was suggested to enable the formative and summative aspects of evaluation to proceed simultaneously within a single study. Finally, by monitoring the effects of environmental factors, the writer was able to begin to distinguish the moderator variables operating at various levels of evaluation and move towards construct validation of a theoretical model of training and evaluation.