Potassium (K) deficiency in cotton (Gossypium hirsutum L.) grown on soils with adequate soil test K levels has become more prevalent. Field experiments were conducted with or without irrigation as the main plot and K fertilizer rates as subplot treatments. Granular K was applied preseason at rates of 0, 45, 90, 134, 179, and 224 kg K ha(-1). Two in-season foliar K programs applied a total of either 2.8 or 5.6 kg K ha(-1) were included to compare against the check. Experiments were set up on two distinct soil mapping units (SMUs) in Starkville, MS, including a Leeper silty clay loam (LSCL; fine, smectitic, nonacid, thermic Vertic Epiaquepts) and a Marietta fine sandy loam (MFSL; fine-loamy, siliceous, active, thermic Fluvaquentic Eutrudepts). Response variables collected include lint yield and fiber quality, leaf K concentrations, and cotton growth and development parameters. Across all years of the experiment, LSCL lint yields increased at a rate of 41 and 118 kg ha(-1) per 45 kg K ha(-1) under rainfed and irrigated conditions, respectively. On the MFSL, lint yield increase was not significant for either rainfed or irrigated conditions. Leaf K concentrations at first bloom and 4 weeks after first bloom responded quadratically to K rate, with peak leaf K concentrations occurring between 134 and 179 kg K ha(-1). Differences in cotton growth parameters were observed; however, differences were likely physiologically insignificant. For all parameters, foliar K application programs did not differ from the untreated check. This investigation showed that foliar K application alone are insufficient, and responses to granular K and irrigation were dependent on SMUs.