Temporal and spatial availability of K can influence soybean [Glycine max (L.) Merr.] productivity. This study quantified the impact of initial soil K concentrations, soil-water content, and soybean K uptake on soil K pools (water-extractable solution-phase K [Ksol], 1 mol L-1 NH4OAc-extractable K, referred to as exchangeable K (K exch), and 5-min sodium tetraphenylboron [NaTPB] extractable K [KTPB]) and compared predictions of plant K availability using the NaTPB vs. NH4OAc tests. Soil and soybean samples were collected five times between the VE and R6 development stages in 2003 and 2004 from fallow and cropped no-till Toronto-Millbrook silt loam Alfisols at the Throckmorton Purdue Agricultural Center. Gravimetric soil water content was measured weekly during the growing season. Low-, medium-, and high-K fertility plots were replicated four times. Initial Kexch levels ranged from 60 to 290 and from 50 to 90 mg kg-1 at the 0- to 10- and 10- to 20-cm depths, respectively. Medium- and high-fertility soils had the highest grain yields. The 0- to 5-cm soil layer had the highest Kexch levels, water availability, and soybean K uptake. Compared with Kexch, Ksol levels were less stratified and the surface layer was less dynamic. The NaTPB extraction was a better predictor of soybean K uptake in 1 yr, but across both years, NH 4OAc was superior. Potassium measurements in the 0- to 10-cm soil layer provided slightly better estimates of plant K uptake than those in the 0- to 20-cm layer. Greater water availability and K uptake in the 0- to 5-cm soil layer suggest that under evenly distributed, intermittent rainfall conditions in somewhat poorly drained soils, vertical soil K stratification might not be a concern for no-till soybean production.