Developments over the past 15 years in our understanding of streamflow generation processes are reviewed. Examples from small- and intermediate-sized catchments in a wide range of hydrogeologic, climatic, and vegetative environments in New Zealand and northeastern Australia are used to illustrate the development and present state of knowledge. Increasingly detailed studies of small catchments and of hillslope and valley bottom segments have revealed increasingly complex and variable responses both in time and space. Although these studies have greatly increased conceptual understanding of flow generation, there has not yet been a corresponding increase in ability to model or predict flow processes. Further progress will require integration of field studies with chemical, physical, and mathematical model studies of specific processes to test hypotheses derived from previous work at well-documented field sites. Concentration of the necessary skills and studies at a few intensively studied locations is also required, but this will impose new and difficult challenges for cooperation among specialists, research organizations, and funding sources.