Abstract: Current grading of mold contamination mainly relies on sensory evaluation in tobacco industry, which leads to inefficient assessment and management of moldy tobacco with significant discrepancies. To develop an objective and effective grading system for mold contamination, this study selected a mixed fungal strain of Aspergillus niger and Penicillium citrinum as experimental inoculum, based on the predominant cultivable microorganisms in mold-contaminated tobacco leaves. Under controlled temperature and humidity conditions, tobacco strips were inoculated with the mixed spore suspension, and the progression of mold growth was monitored by measuring the mycelial distribution area, mold colony count, and CO₂ concentration. The results revealed a power-law correlation between the proportion of mycelial distribution area and mold count per gram, which could serve as a reliable quantitative indicator of mold severity and progression. Furthermore, the inflection point in CO₂ concentration during early-stage mold growth corresponded to a specific threshold in mycelial distribution area, marking the onset of the logarithmic growth phase. Using K-means clustering analysis, the degree of mold contamination beyond this inflection point was classified into five distinct severity levels based on the proportion of mycelial distribution area. As validated against mold colony counts, this grading method exhibited an average misgrading rate of only 1.16%. The grading system can increase the evaluation accurancy for mold contamination in tobacco, and propose a more precise and efficient tool for quality control in tobacco storage and management.