Extensive research has been conducted to understand the relationships between K fertilization, grain yield, and soil-test K but more research is needed to better understand K cycling and high temporal soil-test K variation. The objective of this study was to determine aboveground K uptake by corn and soybean, K removal with grain harvest, and K recycling to the soil. Soil and plant samples were collected two years from 13 Iowa field trials. Two trials were managed with continuous corn and 11 trials were managed with corn-soybean rotations. The continuous corn trials included four N-K rate combinations, and the corn-soybean rotation trials included five K rates. Plant samples taken from all treatments included plant and grain at physiological maturity and grain at harvest time. Additionally, crop residue samples were taken from two contrasting K fertilizer treatments (no K or up to 225 kg K ha^-1) at roughly 45-day intervals from harvest until spring (early April). Fertilizer effects on grain yield and K removal were observed in soils testing below 170 mg K kg^-1 and grain K concentration seldom was affected. However, K increased total plant K uptake at the physiological maturity growth stage at most sites. The largest soybean K loss occurred between physiological maturity and grain harvest, but for corn it was between physiological maturity and harvest or between harvest and late fall (before snowfall or soils froze) depending on rainfall. There was no K loss during winter, and there was an additional small loss in early spring as rainfall increased. The K application rate had no clear effect on patterns of plant K loss. In most years and sites, less than 10% of the K uptake in soybean vegetative plant parts at the physiological maturity stage (except grain) remained in the residue by spring. For corn results were more variable, however, and 30 to 50% of the K uptake remained in the residue by spring. The larger variation for corn was explained mainly by a larger effect of variation in rainfall than for soybean. The rate of plant and residue K loss to the soil before and after grain harvest partly explain usually high post-harvest soil-test K temporal variability.