The A and B horizons of 28 Delaware soils were incubated with 0, 200, 1000 and 2000 ug K/g of soil for one week, one, two and six months. Potassium soil test values were determined using the double acid extractant. The amount of K necessary to raise the soil test value one unit varied between five and 12 ug K/g of soil and was influenced by time of incubation, soil type, amount of K added and soil horizon. Correlation coefficients between the units of K necessary to raise the soil test values one unit and sand, silt and clay and cation exchange capacity are presented. Increasing soil test values to some predetermined point does not appear agronomically or economically sound for the Coastal plain soils of Delaware. Using this procedure to determine fertilizer K recommendations results in recommendations several fold higher than those recommendations which result from using field crop response to determine the amount of K fertilizer to be applied.