Arrowleaf clover (Trifolium vesiculosum Savi ‘Yuchi’) is a winter annual legume that can be used to extend the grazing of warm-season perennial grasses in forage production systems. Stand establishment from overseeding often fails because of unfavorable climatic conditions and nutrient deficiencies in soils. The objectives of this research were to determine the effects of P and K fertilization on yield and nutrient uptake of arrowleaf clover and accumulation of soil P and K. Arrowleaf clover was seeded in field plots in the fall of 1978. Fertilizer treatments consisted of an incomplete factorial arrangement of multiple rates of P, K, and P plus K. Rates of P and K included a control that did not receive fertilizer; 20, 40, and 80 kg P ha−1; 37, 74, and 148 kg K ha−1; and P-K combinations of 20:37, 40:74, and 80:148 kg ha−1. The soil type was a Taloka silt loam (Mollic Albaqualfs). Yield and N, P, and K uptake in forage were determined in four consecutive years starting in 1979. Soil test indices were measured prior to seeding in 1977, in the fall of 1978, and after four annual fertilizations in 1982. Regression equations with linear, quadratic, and interaction terms were calculated to describe dependent variable response surfaces for each year and the average over years. Phosphorus fertilization increased yield each year. Yield was not increased with K fertilization alone, but there was a positive P × K interaction in the 1982 season indicating soil K levels were decreasing. Predicted maximum yields were 6561, 6490, 6516, and 6860 kg ha−1 due to applications of 50.8, 52.4, 54.0 and 57.2 kg P ha−1, respectively, at fixed rates of 0, 37, 74, and 148 kg K ha−1. Phosphorus increased N, P, and K uptake and when applied in combinations with K, some interactions occurred. The soil test index for P increased slowly (0.5 kg ha−1 yr−1) with P rates of 20 kg ha−1 annually; however, a rapid accumulation of soil P occurred with four annual applications of 80 kg P ha−1 (16.5 kg ha−1 yr−1). Approximately 107, 16, 6, and 3 annual applications of 20, 40, 60, and 80 kg P ha−1, respectively, would be required to increase soil test values for P to a sufficient level. Soil test values for K were increased from 133 to 308 kg ha−1 with four annual applications of 148 kg K ha−1. However, after four annual applications of 80 kg P ha−1 in combination with 148 kg K ha−1, the soil test value for K was only increased to 192 kg ha−1 due to the marked response from P fertilization and subsequent increase of K removal in forage.