on soils having Mehlich-1 extractable K of 177 kg ha Ϫ1 or less. Yield response to foliar K continued through Foliar applications of K may be used to supplement soil applications two years of soil-applied 112 kg K ha Ϫ1 plus two years to maximize yields of cotton (Gossypium hirsutum L.). Response to foliar K applications may be improved by choice of K source, buffering of foliar-applied K (four applications per year at 4.1 kg the spray solution, or applying K with B. Research was conducted ha Ϫ1 each). Oosterhuis (1993) indicated that responses on a Collins silt loam (coarse-silty, mixed, acid, thermic Aquic Udifluto foliar K can probably be expected when Mehlich-3 vents) and on a Memphis silt loam (fine-silty, mixed, active, thermic soil K level is 308 kg K ha Ϫ1 or lower. This extractable Typic Hapludalfs) to evaluate KNO 3 , K 2 SO 4 , K 2 S 2 O 3 , and KCl as K K level is similar to the 177 level reported above, using sources. A second study evaluated foliar KNO 3 and K 2 SO 4 solutions Mehlich-1 extractant, since Mehlich 3 extracts approxiapplied unbuffered and buffered to pH 6 and 4 on cotton K nutrition mately 1.5 times more K than Mehlich 1 (H.G. Savoy, and yield. Foliar K in both studies was applied at 4.1 kg K ha Ϫ1 per personal communication, 1998). Tillage system may also application. A third study evaluated combinations of soil-applied and contribute to the response to K fertilization. Howard foliar-applied B and K. Foliar treatments were applied in 93.5 L ha Ϫ1 et al. (1997) reported that no-tillage cotton yields were water at early flower or 2 wk after and repeated on a 9-to 14-d interval between the four applications. Yields from the four K sources increased by applying K to a soil having 225 kg Mehaveraged 10% higher than the untreated check and yields with KNO 3 lich-1 extractable K ha Ϫ1 , but conventional-till yields were 4% higher than the other K sources. Buffering two K source were not increased by applying K to a soil having 193 solutions to pH 4 resulted in 10% higher yields than the check or kg extractable K ha Ϫ1. unbuffered K solutions. Adding a surfactant (ethoxylated alkyl aryl The foliar-applied K source may also affect yield rephosphate esters) to KNO 3 resulted in 5% higher yields then the sponse to foliar fertilization. Miley and Oosterhuis check. Compared with untreated check yields, soil-applied B at 0.56 (1994) summarized three years of evaluating KNO 3 , kg B ha Ϫ1 increased yields by 6%, four foliar applications of 0.11 kg K 2 SO 4 , K 2 S 2 O 3 , KCl, and K 2 CO 3 as foliar K sources and B ha Ϫ1 increased yields by 8%, and four foliar applications of 0.11 reported that KNO 3 increased lint yields, relative to the kg B plus 4.1 kg K ha Ϫ1 increased yields by 13%. Foliar K solution other sources, in two of the three years. Mullins and buffering and/or the inclusion of foliar B are relatively inexpensive ways of improving yield response. Based on yield increases in this Burmester (1995) reported that lint yields were instudy, these treatments should return 8 to 10 times the pr...