Phosphorus (P) and potassium (K) fertilization costs, yields, and economic returns associated with various P and K fertilization levels were evaluated on corn and soybeans in rotation in northeastern Iowa from 1979 to 1989. The treatments were 0, 20, and 40 lb P/acre and 0, 60, and 120 lb K/acre, in all nine combinations, plus the high rate (40+120) applied on alternate years. (The latter treatment applied, respectively, in odd or even years was averaged into one treatment.) The initial soil test levels averaged 57 pounds of P and 355 pounds ofK. The P-K treatments did not significantly explain the variation in corn or soybean yields. Net returns were found directly related to the cost of the treatment, so that the control treatment (0+0) achieved the highest returns. The cost of using the 20+60 treatment to maintain soil test levels in the high to very high range was $24/acre per year. Annual application of 40+120 cost $45/acre. A sufficiency approach to applying P and K could drastically reduce fertilizer costs for high-testing soils. Further research is needed to determine if recommendations from this approach can be lowered in some cases. Extension workers must continue to develop ways to aid farmers in realizing the usefulness of soil testing. Farmers must carefully evaluate their fertilization needs in conjunction with soil test results. Together, a more sustainable approach to P and K fertilization for corn and soybeans may be attained.