This is a report on four experiments on foliar fertilization of cotton with K conducted in Arkansas in 1991. Irrigation was employed in all but experiment 4. Experiments 1 and 2 were studies on the leaf-burn effects of various rates and concentrations of potassium nitrate (KNO₃) alone and in combination with other fertilizers. Experiments 3 and 4 compared the effects of foliar applications of KNO₃ and other potassium (K) sources on cotton yield and other parameters. In experiments 1 and 2, no serious leaf-burn symptoms resulted from foliar applications of KNO₃ at a rate of 10 lb in 5, 7.5, and 10 gal/A and at a rate of 20 lb in 10, 15, and 20 gal/A and from foliar applications of 5 or 10 gal/A of undiluted liquid mixed fertilizers containing KNO₃ in combination with urea and boron with the following percentages of N, P₂O₅, K₂O and B: 13-0-6-0.2, 10-0-10-0.3, 15-0-8-0.3, and 3-0-11-0.3. In experiment 3, no significant effects on lint yield, boll weight, petiole K and petiole NO₃-N resulted from four biweekly foliar applications of five salts of potassium including nitrate, sulfate, chloride, thiosulfate and carbonate when applied at a rate of 10 lb KNO₃ or the K equivalent thereof in 10 gal/A of solution. In the nonirrigated experiment, no. 4, cotton yields were significantly increased by the following foliar treatments: 1.3 lb N/A and 4.4 lb K₂O/A as both KNO₃ and KCl plus urea, 4.5 lb N/A as urea, and 0.27-0.45-0.96 as a commercial liquid mixed grade. There were no significant differences among the yields produced by these treatments, the important economic result being that KCl, the lowest-cost source per lb of K₂O, performed as well as KNO₃ in increasing cotton yield. These results, however, should be considered with caution, pending further research.