The effect of spatial-scale variations on water harvesting has been evaluated at micro-catchment, hillside/farm, and watershed (=catchment) scales in three relatively dry environments in Syria, Pakistan and Egypt. Micro-catchment water harvesting captures localized runoff only through independent micro-catchment systems and is not influenced by hill slope runoff and stream flows. In Syria, it was found that only a fraction of total runoff from a catchment is collected, with no significant effect on water supply downstream. Therefore, it is less sensitive to spatial-scale variations (rainfall, topography, soils) and it is less likely to create upstream-downstream water-use conflicts. Micro-catchment water harvesting established fodder shrubs, with 3-7 fold growth improvement and 27 to 90% survival rate. In contrast, spatial rainfall and geo-hydrological variations greatly influenced the potential of water harvesting at the watershed scale in north-west Egypt. The study revealed the importance of scientific data and appropriate assessment methods. Due to potential socio-economic disputes, the institutional arrangements for water allocation and conflict resolution are critical at this scale. At the farm scale, the low-cost farm runoff structures (US$ 48 per structure), effectively regulated the runoff, reduced terrace damage and improved field soil moisture, and were shown to be suitable for water harvesting on terraces in Barani, Pakistan. The study infers that water harvesting in the drylands is a viable option to improve productivity and conserve natural resources, if it is appropriately implemented. Nevertheless, the size of the water harvesting system and the spatial scale at which it is to be implemented determine the options for water harvesting.