Watermelon [Citrullus lanatus (Thunb.) Matsum. & Nakai] were grown with three rates each of lime, gypsum, and K during two seasons to evaluate their effects on fruit production and mineral concentration. The first experimental site was a recently cleared Sparr fine sand with an initial water pH of 5.0 and Mehlich I extractable K of 8 mg·kg-1 (very low) and 20 mg·kg-1 Ca (very low). The second site was a virgin Pomona fine sand with a water pH of 4.8, 28 mg·kg-1 K (low), and 612 mg·kg-1 Ca (high). `Crimson Sweet' fruit yields were reduced 10% with an increase in lime rate from 0 to 4.48 t·ha-1 in the first season. In the second season, lime rate had no effect on yield. In both seasons, fruit yields were reduced 14% with an increase in Ca from gypsum from 0 to 1.12 t·ha-1. On the soil testing very low in K, yield increased with an increase in K rate from 90 to 224 kg·ha-1 with no lime or gypsum. On the soil testing low in K, greatest yields were obtained with 90 kg·ha-1 K with no lime and gypsum. Application of lime and gypsum increased Ca and decreased K in seedlings but not consistently in older leaf and fruit tissues. An increase in K application increased leaf K in the first season but not in the second. Fruit firmness and soluble solids content were not consistently affected by treatment during the two seasons. Thus, on soils low in toxic elements (Mn and Al) such as used in this study, watermelon will grow well and tolerate a wide range of soil pH values without additional Ca from lime or gypsum.