At global scale nutrient flows are unbalanced, resulting in huge surpluses in Western Europe versus negative balances in Africa. Existing policy and socio-economic environments in different parts of the world are the major cause of this situation. At lower spatial scales, a much more complex and diverse picture emerges. In Sub-Saharan Africa, different levels and causes of soil fertility decline are found among agro-ecological zones, soil types, farm systems, wealth groups, urban–rural households, cash and food crops, home fields and bush fields. Technology development has been the traditional response to address nutrient imbalances in general, and soil fertility decline in Africa in particular. Farm households have continued to develop and adapt existing technology to changing situations. National and international research institutions have followed a variety of changing approaches of which the recently developed participatory approaches have yielded some impressive results in isolated cases. These efforts have, however, not led to the necessary massive reversion of the trend in soil fertility decline. The Dutch policies on nutrient use and the Indonesian policy to adopt Integrated Pest Management are two examples, associated with such major trend reversions. This suggests that promoting and supporting participatory technologies have limited impact when no attention is paid to participatory policy development and implementation. In order to mobilise farm households in a trend reversion, a number of conditions should be met such as stable prices for agricultural outputs, better input/output prices ratios, influence of land users on the research agenda and private-public initiatives focused on smallholders. This observation calls for the establishment of interactive landusers-science-policy triangles at various scales (local, national and international) in which joint learning and mediating may lead to more informed decision making, more focused design of an agricultural sector policy, implementation of policies by effective institutions, and appropriate technology development and implementation. Interventions need to be reoriented: less technology development, more policy influence and institution building.