We developed a new method for analysis of mixed dentition in the lower jaw using only tomosynthesis panoramic radiograph (TSPX) findings obtained from 40 children at the Hellman IIIA stage. For this study, mixed dentition analysis using plaster models and the regression equation of Ono was defined as the A method, while that using plaster models and TSPX findings was defined as the B method. There was no significant difference for the average value of dental arch space between the A and B methods. In addition, we produced simple regression equations (E and F methods) and a multiple regression equation (G method) using the values obtained with TSPX. There were no significant differences for the average value of dental arch space among the A, B, E, F, and G methods. When comparing the mean absolute errors for dental arch space measurement between the B method and the E, F, and G methods using leave-one-out cross validation, the mean absolute error of the dental arch space for the F method was smallest (0.735 mm) among all of the methods. In addition, the mean absolute error of the dental arch space for the G method was 1.055 mm, which was nearly the same for the E method, in which the mesio-distal crown diameters of all permanent teeth and arch length were measured using TSPX. Therefore, we concluded that the present G method is most usable for screening at the chair side, as it requires only 6 measurements for mixed dentition analysis.