For winter wheat (Triticum aestivum L.) that grows during the rainless season, the contribution of groundwater to the root zone (CGWR) is an important water source for growth. Accurately estimating the CGWR is important for making decisions on irrigation and discharge for winter wheat fields and preventing water pollution. Because winter wheat slows and even stops root growth over winter, so the fixed root density distribution function that is suitable for soybean [Glycine max (L.) Merr.] may not suit winter wheat calculations. Therefore, when estimating the CGWR of winter wheat with the numerical model HYDRUS-1D, the root density distribution function should first be determined from two types: fixed or piecewise root density distribution functions. Based on field observations and local weather data for 2004–2005 and 2005–2006, HYDRUS-1D was evaluated with different root density distribution functions by comparing simulated and measured root zone soil water contents. The evaluated model with the most suitable distribution function was used to estimate the daily CGWR for six winter wheat hydrological growth seasons. For all seasons, winter wheat growth was assumed to be at its optimal state. The main results were: (i) a piecewise root density distribution function was the most suitable for winter wheat; (ii) simulated seasonal CGWRs were 154, 128, and 136 mm in the dry, normal, and wet seasons, respectively; and (iii) the CGWR for winter wheat transpiration was about 58, 47, and 69% of the total in dry, normal, and wet seasons, respectively. Overall, we concluded that accurate description of the root density distribution was helpful to estimate the CGWR.