Soil moisture variation is not only affected by climatic conditions but also by soil physical properties and crop physiology as well. The study taking above three factors into consideration and using a simulation model calculates the dynamic to a soil water management project for the vineyards. Golden Muscat and Kyoho vines both pruned in X type and each cultivated in one tenth hectare, were chosen for monitoring and simulating the variations of moisture tension in the soil profiles for one years. While the observed data were simulated and compared, a climatic data model was also used to test its suitability in applying to the simulations. Computer input files included : soil physical properties ; crop physiological data ; climatic data ; numerical conditions ; and irrigation date and depth. In the simulation, a zero tension depth was varied as a parameter. Oncce the zero tension depths were set up, the simulated results were found very close to the observed variation of soil moisture tensions at different layers, suggested the effectiveness of the applicability of the model. No matter the real climatic data or the climatic modelling data was used, the variations of soil moisture tensions at 30 cm and 50 cm depths were well simulated. Their statistical correlation coefficients were greater than 0.739. Using the real rainfall data to replace the simulated precipitations would increase the correlation coefficients up to at least 0.983, indicated that this combination of the input climatic data would give the best predictions.