Effect of long-term fertilization on potassium (K) supplying capacity of paddy and purple soils of China was investigated by employing the quantity/intensity (Q/I) approach. Results show that after long-term fertilization, the treatments of the experiment were found to be in the order of NPKM > NPK > CK > N and NP for both paddy soil and purple soil, in terms of K⁺ equilibrium activity ratio (AR₀^K, in the same order for paddy soil and in the order of NPKM>NPK and CK>N and NP for purple soil in terms of soil labile K (K_L), non-specific K (-ΔK₀) and specific K (K_X). They followed the sequence of NP>N>CK>NPKM and NPK, for purple soil in terms of potential buffering capacity (PBC^K), while PBC^K in paddy soil remained nearly unchanged after 15 years fertilization. The paddy soil was greater in AR₀^K, K_L and-ΔK₀, but lower in PBC^K (21.7 to 29.1) and exchange free energies (-ΔG) (11.4 to 13.9 kJ mol^-1) than the purple soil (PBC^K values of 48.2 to 125.8 and-ΔG values of 14.9 to 19.9 kJ mol^-1). In Treatments N and NP, AR₀^K, K_L,-ΔK₀ and K_X decreased and PBC^K and-ΔG increased, and as a result, soil K depletion was intensified especially in purple soil. The trend of soil K depletion was retarded in Treatments NPK and NPKM. In the experiment, soil K depletion was greater in the purple soil than in the paddy soil, maybe because the K or organic manure application rate was not adequate in the purple soil.