To discuss the interrelation among the water potentials of soil, plant and atmosphere and their relation to evapotranspiration, the water potential of atmosphere, plant and soil were measured and the evapotranspira-tion rate was estimated using a heat balance method. The observations were conducted in wheat fields of a Pseudogley (PSG), an Ordinary Andosol (OAS), and a Sand-Dune Regosol (SRS) in the neighborhood of Sapporo for the period from early June to harvest in 1986,1987 and 1988, respectively. The distribution of roots and the dry weight of the above-ground biomass were also determined. (1) The dry weight of above-ground biomass of wheat was in the order of SRS > PSG > OAS. (2) In PSG, with the least available water capacity of the three fields, the amount of rainfall was low during the vegetative growth and flowering stages. However, the average daily evapotranspiration (ET) had a high value of about 3.3 mm/day, and the daily ratio of latent heat flux to net radiation (IE Rn) was found to be higher than 0.7. In spite of a remarkable decrease in the water potential of atmosphere (PA), plant water potential (PP) was maintained to be high, ranging within — 6 to —1.2 bar. This suggests that water supply from soil to plant roots was satisfied. We ascribed this to the fact that the roots took up the stagnant water directly. This was supported by the following facts 1) lowering of stagnant water level with wheat growth, and 2 ) some wheat roots having touched at the stagnant water pool. (3)The available water capacity of OAS is 5 times greater than that of PSG. The pattern of rainfall was similar as in PSG. However, the average daily ET was a low value of about 2.7 mm.z day, and the lE/Rn ratio varied with growth stage and rainfall event, and has low values, ranging from 0.4 to 0.6 dur-ing a dry spell in the flowering stage. PP decreased remarkably with decreasing PA in this stage, and the value converged into a narrow range of -11 to - 7 bar. These facts suggest that plant might have suffered from water stress, and explain why the dry weight was the least among three fields. (4)In SRS, the available water capacity is 12 times greater than that of PSG. Total amounts of ET during vegetative growth and flowering stages corresponded to total amounts of rainfall in the vegetative stage. Therefore a higher ET of about 3. 65 mm day was found during the dry spell in the flowering stages. The lE/Rn ratios in the growth stages were maintained within a range of 0.6 to 0.8, and also PP values less than ― 8 bar were not found. These facts explain why the good wheat growth resulted.