Soxhlet extraction is a common method of sample preparation. However, there has been no discussion about the efficiency of Soxhlet extraction from different batches and the factors that cause content fluctuation. In this study, Panax ginseng was selected as a model sample. Soxhlet extraction by means of a water bath, which has always been neglected, was identified as a novel key factor in the poor repeat-ability in different batches of Soxhlet extraction, as it can affect the siphon times and reflux time, which have been positively correlated with the ginsenoside contents. By substituting round bottom flasks in the same column, the relative standard deviation of the most fluctuated compound, ginsenoside Rb1, was decreased from 24.6% to 5.02%. Scanning electron microscopy analysis confirmed that the breakdown of the surface of the ginseng powder in the Soxhlet extraction led to a better dissolution of ginsenosides, indicating that chloroform may promote the extraction of ginsenosides by disrupting the cell structure. Moreover, 70% methanol was regarded as the better solvent for extracting the ginsenosides. Overall, this work offers a practical and effective protocol for improving the accuracy and repeatability of Soxhlet extraction methodology for ginsenosides and other analytes.