- The deposition of dew is a common meteorological phenomenon that has been recognized as an important ecosystem element, especially in arid areas. There is some evidence that indicates that there is an increase in the water content of the topsoil during nights in which no dew deposition was observed. The purpose of this study is to describe the daily pattern of changes in water content in the upper soil layers and to identify the mechanism by which water is added to the soil (deposition or direct absorption). Moreover, the gains in soil water content during the night are compared to the dew amounts recorded by the Hiltner balance, and the losses and gains of water in terms of easily measurable environmental parameters are parameterized. Nine 24-h field campaigns took place during the dry season of 2002. During each campaign, the 100-mm topsoil was sampled hourly, and water content at 10-mm increments was obtained. Micrometeorological measurements included incoming and reflected shortwave radiation; net radiation; wind speed at four levels; dry-and wet-bulb temperatures at 1-m height; and soil heat flux. In addition, the changes in mass of an improved microlysimeter were recorded, and dew deposition amounts were measured using a conventional Hiltner dew balance. The results of this study indicate that in the area in which this study was carried out actual dew deposition on a bare soil surface is probably a rare occurrence. There is, however, a clear discernible daily cycle of water content in the upper soil layers. The lack of any evidence of soil surface wetting led to the conclusion that the main process responsible for the observed diurnal change in water content is the direct adsorption of water vapor by the soil. A strong and significant correlation was found between the total adsorption of water vapor by the soil during the period that begins in the early afternoon and ends at sunrise and the total potential evaporation between sunrise and sunset of the previous day. Based on this finding an empirical model is proposed in order to predict the total amount of water adsorbed by the soil during the absorption period. The proposed model is probably site specific but is very simple and easy to implement. An additional outcome of the present study is that, in the area in which it was carried out, artificial condensing plates are poorly correlated to water vapor absorption, and the deposition of dew on them is not indicative of dew deposition on bare soil. © 2004 American Meteorological Society.