**Background:** Potassium deficiency in East Asian soils severely limits rice production. Breeding potassium-efficient rice varieties through QTL mapping requires selecting appropriate parent pairs with contrasting responses to low potassium. Previous studies have used various potassium concentrations (3–16 mg/L K⁺) for screening, but the optimal concentration and ideal parent pair for QTL mapping remain to be established.
**Methods:** Twelve high-yielding rice varieties from East Asia (including japonica types Wuyunjing, 02428, NP, TB309, TQ and indica types 9311, J189, J193, J221, J187, J426, J216) were grown hydroponically at pH 5.8 with ten potassium concentrations (0, 1, 2, 3, 4, 5, 10, 20, 40, 60 mg/L K⁺) for 4 weeks, with nutrient solution replaced every two days. At the five-leaf stage, three replicates per variety per treatment (six plants per replicate) were measured for plant height, root length, root number, fresh root weight, fresh sheath weight, and fresh blade weight. After drying, dry weights and potassium content (via ICP-OES) were measured. Relative values (normalized to 40 mg/L K⁺ treatment) and coefficients of variation (CV%) were calculated. Differences between NP and 9311 were analyzed using paired sample T-tests in IBM SPSS Statistics 23.
**Key Results:** Relative plant height decreased linearly with decreasing potassium concentration across all varieties, with no clear indica-japonica differentiation. Based on consistency and differentiation of relative plant height, fresh sheath weight, and fresh blade weight, NP was identified as low-potassium tolerant and 9311 as low-potassium sensitive. The coefficient of variation for most parameters peaked at 4 mg/L K⁺ treatment, identifying this as the optimal screening concentration. At 4 mg/L K⁺, NP was significantly greater than 9311 in relative root length and fresh sheath weight (p < 0.001), relative fresh blade weight (p < 0.01), and relative plant height (p < 0.05). At 1 mg/L K⁺, 9311 showed typical potassium deficiency symptoms (brown/yellow leaf tips on old leaves) while NP did not. Potassium-related trait analysis showed NP had higher potassium response index (0.0059 vs. 0.0016), potassium absorption rate (0.1742 vs. 0.1328), potassium translocation rate from root to sheath at 1 mg/L K⁺ (0.3773 vs. 0.2493) and at 4 mg/L K⁺ (1.1930 vs. 0.7086), and relative potassium accumulation at 4 mg/L K⁺ (0.1998 vs. 0.1349) compared to 9311. No significant difference in potassium utilization efficiency was found between the two varieties.
**Clinical Implications:** This study identifies NP (japonica, tolerant) and 9311 (indica, sensitive) as a rare parent pair with contrasting indica-japonica backgrounds for mapping potassium-efficient QTLs. The optimal screening concentration of 4 mg/L K⁺ provides a standardized method for future screening. The observed differences in potassium translocation (particularly root-to-sheath transport) suggest that genes controlling SKOR1, NPF, and KT/HAK/KUP transporter families may underlie the differential low-potassium tolerance. Constructing recombinant inbred lines or introgression lines from this parent pair could enable identification of QTLs for high potassium efficiency, offering a breeding strategy to mitigate potassium deficiency in East Asian rice production.