Field trial consisting of seven steps of K fertilization, rate of 2 670 kg K2O/ha, was conducted on K-fixing gleysol in autumn of 1986. For the growing seasons of 1988 and 1989, the uniform rates of potash fertilizer (150 kg K2O/ha) were applied, while for the season of 1990 K fertilization was restored, except of two highest rates. A response of soybeans (season of 1990) was shown in this study. Using 1 000 kg K2O/ha the grain yield increased to 1.07 t/ha or 59%, compared to the control, while trifoliate-leaf composition (in dry matter at the onset of anthesis) was as follows: 0.90% K (symptoms of K deficiency) and 1.67% K as well as 1.94% Mg and 1.27% Mg for the control and 1 000 kg K2O/ha application, respectively. However, an application of the highest K rate in the spring of 1990 resulted in toxic effects on soybeans at the emergence stage (plant density reduction). For this reason, the grain yield increased only by 30% compared to the control. In addition, subsequent effects of K fertilization in the autumn of 1986 (2 670 kg K2O/ha) were found out in the growing season of 1990: the grain yield increased by 49%.