There has been a rapid increase in the availability of improved and short duration varieties, early crop establishment techniques, pest management alternatives, farm machinery and supplemental irrigation. To be useful to Asian rice farmers these new technological components need to be carefully combined to fit the prevailing production environment. This requires a holistic approach to agricultural research that is oriented towards the combination of crop enterprises encountered on, or suitable for, the different land types in rice-growing regions. In formulating such an approach, IRRI's cropping systems programme has tried to avoid complex computational and information processing techniques. Co-operation with representatives from national research organisations in south and southeast Asia (Cropping Systems Working Group, 1975) led to the formulation of a site-related cropping systems research methodology that emphasises decentralised research in the farm community. This methodology is conceptually based on the cropping system as an open, reactive system. The research activities therefore focus on the description and classification of the environment, on the design of improved cropping systems, on their on-farm testing and on methods for the formulation of production programmes.