Soil potassium deficiency has severely reduced the potato field production in China. Fortunately, different potato varieties respond to low-potassium conditions very differently. Therefore, the utilization of potato varieties with low-potassium tolerance is an important approach to reduce potassium application by increasing potassium utilization efficiency, thus promoting the sustainability and the green development of agriculture in China. In this study, 17 biological features of 30 potato varieties were examined under normal potassium (202.5 kg∙hm^-2 K₂O) and low potassium conditions (0 kg∙hm^-2 K₂O), and nine representative features, including leaf area index, root-shoot ratio, shoot dry mass, root dry mass, tuber yield per plant, large tuber yield per plant, small tuber yield per plant, tuber dry mass, and tuber potassium accumulation were selected for subsequent analysis. The results showed that all feature values went down under the low potassium condition. Principal component analysis revealed that the nine features can be transformed into 4 independent comprehensive components, with a cumulative contributive rate of 87.1%. According to the comprehensive evaluation value (D value) and cluster analysis, the 30 varieties can be divided into 6 categories, including seven high tolerant varieties to low potassium: Lucinda, Favorita, Kexin1, Xisen6, Xingjia2, Helan15 and Chuanyin2, and six varieties with moderate tolerance to low potassium: Longshu20, Dingshu3, Jizhang12(W), Jiuen1, Longshu19, Jizhang12(Y). Furthermore, we developed a regression model Y=-0.595+0.247X₅+0.155X₄+0.138X₃+0.167X₈+0.088X₁+0.081X₆+0.097X₉+ 0.053X₂ (R²=0.999, P=0.000) to distinguish the 30 varieties from each other with an accuracy above 90%. In summary, under low potassium condition, several features, including tuber yield per plant, root dry mass, shoot dry mass, tuber dry mass, leaf area index, large tuber yield per plant, tuber potassium accumulation and root-shoot ratio can be used to rapidly identify low potassium tolerant varieties.