In a previous experiment, evaporation from soybeans (Glycine max L.) in two weighing lysimeters with different soil types was found to differ by up to 30%. This occurred despite good canopy development and maintenance of well watered conditions. The present experiment sought to repeat the previous observation and to define its cause. Soybeans were sown in and around the two weighing lysimeters on 9 December 1987 and were well watered through the entire season. The lysimeters, L1 and L2 contained undisturbed blocks of Hanwood loam and Mundiwa clay loam soils, respectively, both Rhodoxeralfs. Crop growth, radiant energy interception, soil heat flux, canopy temperature and root growth were monitored through the season. Plant growth in L2 was slower than in L1 such that by 46 days from sowing (DFS), L1 plants had one leaf more on average than those in L2 and by 76 DFS plants in L2 were about 0.1 m shorter than those in either L1 or in the area immediately surrounding it. The ratio of L2 to L1 daily evaporation was 0.76 during the period 75 to 84 DFS; this being very similar to the effect observed previoulsy. The crop canopy in a 100 m 2 area centred around L2 was reduced in height by removing the top 0.15 m at 85 DFS. This treatment caused the L2:L1 evaporation ratio to increase to 1.07. The effect of reducing the height of plants surrounding L2 was to increase net radiant energy intercepted in the canopy of the L2 plants and to change the turbulent transfer processes over the L2 canopy. Shading from the taller surrounding plants was estimated to have reduced evaporation by 4% while increased aerodynamic resistance above the L2 canopy as the result of the height discontinuity accounted for a further 20% reduction. This study highlights limitations in the application of one dimensional energy balance theory to non-ideal canopy configurations and to the care needed to ensure plant growth within lysimeters is the same as the surrounds.* Visiting scientist and present address: Direccion General de Investigacion y Extension Agrarias, Departamento de Suelos y Riegos Apdo. 240, E-14080 Cordoba, Spain Offprint requests to. W.S. Meyer Differences in soil type and small variations in canopy height within a well watered crop with a closed canopy are not generally expected to cause major spatial variation in evaporation rate. However, in a previous experiment (Meyer et al. 1990) we found evaporation rates from soybeans in weighing lysimeters 50 m apart with different undisturbed soil blocks to be up to 30% different at all times during an irrigation cycle, despite good canopy development. The cause of this difference was not established, but possible effects of soil type on available water could not be sustained while a height discontinuity (approximately 0.1 m) between the shorter crop in one lysimeter and its immediate surrounds may have contributed to its reduced evaporation rate.Experiments by van Bavel et al. (1963) clearly showed that crop height with respect to its surrounds strongly influenced the rate of evaporation. H...