The conventional 1 M NH4OAc-exchangeable potassium (K+) soil test is inadequate in soils where nonexchangeable K+ contributes significantly to crop nutrition. Studies were conducted (i) to compare the abilities of the I M NH4OAc method with a modified NaBPh4 method to estimate critical soil K+ levels, (ii) to estimate the contribution of nonexchangeable K+ to plant-available Iii, (iii) to compare the abilities of the 1 M NH4OAc method and the modified NaBPh4 method to estimate plant dry matter yield and plant-available K+, and (iv) to compare the abilities of both methods to measure soil K+ balance. Winter wheat (Triticum aestivum L. 'Abe') was grown in eleven Midwestern soils in a greenhouse using consecutive 28-d defoliation and regrowth cycles. Soils also were incubated for 6 mo with five K+ rates (0-809 mg K+ kg(-1)). Ammonium acetate- and NaBPh4-extractable K+ (5-min extraction period) were determined in soil samples taken after every three defoliation cycles and after incubation, Critical soil K+ levels could not be determined by either method alone but could be predicted by including cation-exchange capacity (CEC) and illitic K+ content in regression models. Nonexchangeable K+ represented a significant portion of plant-available K+. Plant-available K+ and dry matter (DM) yield were well related to NH4OAc-extractable KI only in soils with low nonexchangeable K+ contribution (r(2) = 0.889 and 0.915, respectively), but they were well related to NaBPh4-extractable KI in all soils (r(2) = 0.984 and 0.874, respectively). Slopes for NH4OAc-extractable K+ vs, soil K+ balance varied widely among soils (0.16-0.68) depending on NH4OAc-extractable K+, illitic K+. and clay content, but for NaBPh4-extractable K+ slopes were near unity, These studies suggest that the modified NaBPh4 method may be a superior Kt soil test compared to the NH4OAc method. Illite content and CEC data may help in developing better soil Kt management guidelines.