Objectives The purpose of this study is to predict the initial average weight loss in overweight and obese adults taking Gamitaeeumjowi-tang. Methods A retrospective analysis was performed on 25,988 overweight and obese people taking Gamitaeeumjowi-tang from January, 2017 to May, 2019. Exploratory data analysis (EDA) was performed using age, gender, weight, BMI (body mass index), Lf/Hf of heart rate variability (HRV), and patient health questionnaire-9 (PHQ-9). After categorizing using significant variables, multiple regression analysis was conducted with the average weight loss corresponding to the combinations of categories as the dependent variable. Results Through EDA, age, gender, initial weight, and initial BMI were identified as significant variables for the initial weight loss. As a result of multiple regression analysis after categorizing the significant variables, the R2 was 0.820 and the Adj. R2 was 0.798. Conclusions In this study, multiple regression analysis was performed using the combination of categorized age, sex, initial weight, and initial BMI. As a result, the initial average weight loss could be predicted with high explanatory power.