peer reviewed ; The project used different methods of data collection: tests, questionnaires and systematic observation of pupils and teachers. In developing a multiple viewpoint approach, we attenpted to sketch a detached view of the implementation of DPE and of tis results. Goals were as follows: (1) Exploring the possible effects of the programme in various areas of children's life that do not arise directly from their activities during physical education lessons. (2) Identifying children's school-life context, in order to identify factors influencing several effects of DPE. The various viewports considered, the results, and the knowledge of external variables all allowed us to identify a convergent package of indications, thus allowing us to think in terms of causality. Often, the benefits came from the children's involvement in the experimental programme. Improvements of motor performance in primary school age children arise from a maturation process and from practice and learning. Nevertheless, it is particularly hard to identify the relative contributions of these two sources. Tracing pupil's development, fro the starting level at beginning of the school year, appeared to allow isolation of the part played by children's activities whitin the school environement. In all grade levels, comparisons confirmed pilot classes superiority. Performances were significantly better in many tests. However, during the year, it seemed that the development of performance due to maturation masked the potential effects of physical activities. A decline in favourable attitudes was observed in every school at the beginning and the end of the school year. A slide from very favourable attitudes towards simply favourable attitudes was observed. . ; Un programme d'éducaton physique quotidienne (leçons de 40 minutes) fut mis en oeuvre dans 14 écoles (maternelles et primaires) pendant trois ans. Ces écoles étaient situées dans la Communauté française de Belgique. Plus de 3.500 enfants furent impliqués dans le programme. Celui-ci ...