A field test was conducted in 1992 and 1993 on a Lucedale sandy clay loam (Rhodic Paleudults) in central Alabama with a medium soil test rating for K. The primary objective of the study was to evaluate cotton response to the rate of soil applied K and the source of foliar K. Treatments consisted of rates of soil applied K with and without foliar applications of K. Soil K was applied preplant at rates of 0, 30 and 60 lb K₂O A^-1. Sources of foliar K in 1992 included K nitrate (KNO₃: 14-0-44), K chloride (KCl: 0-0-62) and K thiosulfate (K(2)S(2)O₃: 0-0-25-17S, liquid). In 1993, K sulfate (K(2)SO₃: 0-0-49-17S) was included as an additional source. During both years a sulfur check was included which received S as magnesium sulfate (MgSO₃: 12.9% S) at a rate (3.1 lb A^-1) equivalent to the S applied as K thiosulfate. Foliar N was applied to all treatments at a rate of 1.38 lb A^-1 as urea or KNO₃. For all treatments, the respective fertilizer sources were dissolved in water and diluted to give a total application rate of 10 gal A^-1. The rate of foliar K was 4.4 lb K₂O A^-1 per application. Foliar K treatments were initiated at ~ 2 weeks after first white flower and repeated at 10-14 day intervals for a total of 4 applications. All treatments received a total of 5.12 lb of foliar N A^-1 and all of the foliar K treatments received a total of 17.6 lb of foliar K₂O A^-1. A response to treatments was observed only in 1992. In 1992, a yield response was obtained to the rate of soil K. A preplant soil application of 60 lb K₂O A^-1 increased lint yields by 187 lb A^-1. Foliar treatments consistently increased yields but the differences were not significant. There were no differences among the foliar K sources tested. Lint quality as measured by HVI was not affected by any treatments during either year of the test. In 1992, petiole K concentrations were increased by the rate of soil K, but they were not affected by foliar applications of K.