The complex permittivity of baijiu varies with frequency, and dielectric spectroscopy has been used to evaluate the quality. To simplify the analysis and reduce the number of the parameters, a dielectric relaxation model is often used to fit the permittivity data. However, existing fitting methods such as the least squares and particle swarm optimization methods are often computationally complex and require preset initial values. Therefore, a simpler calculation method of the relaxation parameters considering the geometric characteristics of the permittivity spectrum is proposed. It is based on the relationship between the Cole-Cole relaxation parameters and the Cole-Cole diagram, which is fitted by a geometric method. First, the concepts of the Cole-Cole parameters and the diagram are introduced, and then the process of obtaining the parameters from the complex permittivity measurement data is explained. Taking baijiu with 56% alcohol by volume (ABV) as an example, the fitting is better than the least squares method and similar to the particle swarm optimization. This method is then used for the parameter fitting of baijiu with ABV of 42-52%, and the average error is less than 1%, demonstrating its wider applicability. Finally, a prediction model is used for baijiu with 53% ABV, and the error is only 1.51%. Hence, the method can be applied to the measurement of ABV of baijiu.