Field and laboratory studies were conducted to evaluate soil test and plant analysis criteria that form the basis for potassium (K) fertilization for cotton. Field trials were located in Arkansas at the Cotton Branch Experiment Station (CBES) on irrigated Memphis, Grenada and Calloway silt loams and in Craighead County (CC) on non-irrigated Amagon silt loam and Beulah, Dundee and Mhoon fine sandy loams. Potassium rates of 0, 30, 60, 90, 120, and 150 lb K₂O/A were applied at both sites. Yield response was compared to normal neutral ammonium acetetate extractable K in soils and 2% acetic acid extractable K in petioles.

Yield responses to applied K was dependent on both available soil K and yield potential. At the CBES location, soil test K was 220 lb/A (medium); and response occurred only when yields exceeded 1,100 lb lint/A. At the CC location, soil test K was 140 lb/A (low); and response occurred when yields exceeded 500 lb lint/A.

Petiole analyses were reliable criteria for determining K deficiency in cotton when water was adequate and yields were 1.5 bales/A or more.