Most methods of measuring soil-surface evaporation originate from those of measuring water-surface evaporation, and hence they are not necessarily suitable for making measurements of evaporation from dry land surfaces. The major reason is that, as observation shows, when the dry surface layer (DSL) forms on the soil, the constant-flux layer of water vapor is barely developed in the surface air layer. However, during most of the day, the upward water vapor flux in the DSL is nearly constant with depth. The principle of the DSL method is to estimate this upward near-constant flux in the DSL from the depth profiles of temperature and soil water content in the layer. The DSL method was applied to measurements of evaporation from a sand surface at the HEIFE desert station and they were compared to the decreases in soil moisture in the top 1 m of sand, from which the following results were obtained. (a) Estimates of daily evaporation about a week after rainfall of 15 mm or so, which were made on the basis of seven measurements taken by the DSL method during the day, were in good agreement with the daily decreases in soil moisture in the top 30 cm of sand. (b) Daily evaporation from dry soil can be approximately estimated from one measurement of evaporation rate made by the DSL method in the afternoon and the time lapsed since the last rainfall.