Soil is the most important source of K for plants; many studies have focused on soil K status and phytoavailability. However, there is still a lack of quantitative explanation for how soils hold plant‐available K. Based on the previous study, the ability to hold plant‐available K for soil was defined as soil potassium‐holding capacity (SKHC). Therefore, this study aimed to investigate the mechanisms underlying SKHC. Fourteen soils from different sites in China were collected and the contribution of soil organic matter (SOM) and clay minerals to SKHC were analyzed based on waterlogging simulation experiments, chemical determination, and data‐fitting analysis. Results showed that SKHC mainly depended on the absorption and holding effects of exchangeable sites deriving from both SOM and clay minerals (presented as cation exchange capacity [CEC]) and nonexchangeable sites deriving from clay minerals, which contributed 14.7 and 82.5%, respectively, to SKHC, at similar soil water‐soluble K levels. After removing the SOM from the soil, the average SOM and CEC decreased from 26.0 to 7.73 g kg–1 and from 13.7 to 11.8 cmol kg–1, respectively. The SOM contributed an average of 24.5% to CEC, with the remaining 75.5% provided by clay minerals. Further analysis showed when SOM increased by 1%, CEC increased by 1.14 cmol kg–1, a relative proportion of 8.32% of the initial soil. Therefore, SKHC is mainly determined by the nonexchangeable sites of soil clay minerals, and increasing CEC by improving SOM content to enhance SKHC should be an effective but slow approach.