Tomato(Lycopersicon esculentumMill. cv. Sunny) was grown with the full‐bed polyethylene mulch‐seepage (modified furrow) irrigation system for three seasons to evaluate the effects of potassium (K) sources and K rates on fruit yields and leaf K concentrations. Soil in the experimental area was an Eau Gallie fine sand (sandy, siliceous, hyperthermic Alfic Haplaquods) that varied from 12 mg kg‐1(very low) to 56 mg kg‐1(medium) K by Mehlich I extraction prior to planting. Potassium sources, potassium chloride (KC1), potassium nitrate (KNO3), and potassium sulfate (K2SO4) were evaluated at 0, 90, 180, 270, and 360 kg kg‐1K rates. Nitrogen (N) was applied at 270 kg kg‐1and P at 43 kg kg‐1with all K rates. Yields of extra‐large and marketable total yields in one season were higher (P<0.05) with potassium nitrate (KNO3) than with KC1. Maximum yields were produced from 270 to 360 kg K kg‐1, regardless of pre‐plant soil K concentrations. In the shoots, K concentrations increased with increasing K rates. At 90 and 180 kg K kg‐1application rates, K concentrations in the shoots were <2% and the plants and fruits had K‐deficiency symptoms.