The objectives of this research were: (1) to evaluate the influence of three initial P levels created artificially and of four annual P applications on the yields and P content of the corn-soybean rotation, (2) to measure the evolution of the available soil P content as a consequence of the initial P levels and annual P applications, and (3) to assess in economic terms the influence of the initial P levels and annual P applications on corn and soybean yields and on the soil P content;Experimental data used in this study came from a field experiment conducted in central Iowa during the period 1974-1982. In the spring of 1974, applications of 0, 147, and 294 kg P per ha were applied to the whole plots to create three initial levels of soil P. After the first year, annual applications of 0, 11, 22, and 33 kg per ha were applied to split plots. Corn and soybeans were grown each year in rotation;The initial P levels and annual P applications did not consistently affect crop yields, but year-to-year variability in yield resulted from differences in moisture stress. Phosphorus content in corn leaves, P content in the grain, P removed by the grain, and available soil P were affected positively by P applications;Available soil P generally decreased over time. The rate of decrease was increased by an increase in the level of P initially applied, decreased by the rate of annual P application, and decreased over years;Because the treatments had inconsistent effects on yields, it was impossible to obtain a good economic analysis. To maintain a desired soil P level, the annual application of P had to increase as the desired available soil P level increased and/or the initial P application decreased. An inexpensive way to obtain any level of available soil P was to apply enough P to raise the soil P level and then maintain that level by annual P applications.