Little is known about the combined effect of agriculture and climate on soil aggregate stability. Few places in the world allow the study of the in¯uences of both factors. Climatological gradients in Bolivia, where the agriculture is historically developed at different altitudes, provided opportunity to study the effects of climate and agriculture on soil aggregate stability. Aggregate stability was measured by the modi®ed Emerson water dispersion test (MEWDT), drop impacts (CND and TDI) and ultrasonic disruption (UD) on agriculture (Zea mays L.) and scrubland (Acacia caven (Mol.) Mol., Leguminosae) land uses. Different methods and tests were applied in order to validate their ability to characterise aggregate stability. Each test had different range of energy (MEWD TD CN UD) and thus they measured different aggregate resistance. The results showed that soil erodibility was greater under agricultural land use than under scrubland. These differences were greater at the arid site, where the scrubland soils had aggregates nine times more stable (test TDI), compared to the agricultural soil. For both lands uses, climate in¯uenced soil aggregate stability. The higher the mean annual rainfall the greater was the soil aggregate stability. For the CND test, aggregates were between 2 and 4 times more stable on areas with 900±1000 mm yr À1 than on 300±400 mm yr À1 . Organic matter was the key factor determining soil aggregate stability on agriculture land (r 2 ˆ 0X80) and scrubland (r 2 ˆ 0X78) use. Agriculture was a driving force leading to the degradation of soil structure. This re¯ects that land use can exert a great in¯uence on soil structure. Moreover, if climate change reduces mean annual rainfall, then aggregate stability will be reduced drastically. # 2000 Elsevier Science B.V. All rights reserved.