Nine field studies were conducted over a 3-yr period to determine the response of carrot (Daucus carota L.) to soilapplied K on sandy to loamy sand Orthic Podzol soils in Prince Edward Island (PEI), Canada. Mid-season soil K levels ranged from 34 to 103 µg g-1, and sites were classified as L– (very low), L (low) and M (medium) (Mehlich 3 extractable) according to the PEI Soil and Feed Testing Laboratory. Cultivars used were Imperator-type and the crops were managed to conform to local growing practices. Treatments consisted of preplant broadcast applied muriate of potash (KCl) at 0, 75, 150, 225 or 300 kg K ha-1. When no K was applied, total carrot yields on the L–, L and M soil K ratings were 70, 90 and 98% of the maximum yield, respectively. To achieve 95% of the maximum yield, 150, 75 and 0 kg K ha-1 were required on the L–, L and M soils, respectively. Increasing rates of applied K linearly increased the K content of carrot petioles and soil samples taken at harvest. Potassium content in the carrot roots increased linearly with the rate of applied K, but parallel line analyses indicated that separate K response lines were required for each K soil test rating. Rates of 300, 150 and 75 kg K ha-1 were required to increase root K content to above the published standard of 27 mg g-1 (DM basis) on the L–, L, and M soil K sites, respectively. Key words: Carrot, potassium, soil K rating, nutrient content, yield