In considering the developing world it serves little purpose to link land administration to geospatial data infrastructures (GDIs) without taking into account land management. A wider, institutionally based focus is necessary to analyse the multiple local initiatives and large-scale informal land management/administration being undertaken. A focus on land management/administration organisations linked to the cadastre is necessary but not sufficient. There is a wide range of other institutions undertaking land management/administration each with their own form of LIS/GIS, using a range of spatial units. Consequently, the definition of GDI for a developing country should differ from that of a developed country's GDI. It should be institutionally based, and accommodate multiple institutions and a range of spatial units. In terms of this thinking, a range of research questions are posed about inter-operability, accuracy and coordinates, legal and regulatory frameworks, cost and affordability, and appropriate GDIs for conditions in the developing world. References and bibliography [Taken from journal headnote].