Fertilizer P provides three benefits to subsequent crops. The first is that P not lost in produce or leachate remains in soil. However, its effectiveness usually decreases with time because the P slowly diffuses into the adsorbing particle. The second benefit stems from the resulting increased negative charge and decreased buffering capacity. We test whether this decreases the P required for a given yield. The third benefit occurs because, given enough P, the penetration reaction stops. We test whether it is then sufficient to only supply P needed by the plants. We incubated for 35 d at 60°C samples of a P-deficient soil to which we had added five levels of P. At the end of this incubation period, we measured P sorption and desorption on one set of samples of the soil. On another set, we added several different levels of P and measured plant response. On a third set, we similarly added several different levels of P but incubated soil plus the extra P for a further 3 d at 60°C before measuring plant response. Incubating the soil with P decreased the P buffering capacity and increased the effectiveness of P applications. The effects were nonlinear with greatest effect occurring at low levels of application. The incubation with P also virtually eliminated sorption-desorption hysteresis. With increasing levels of incubated P, the difference in plant response between the once-incubated and the twice-incubated treatments decreased linearly. All of the observations are consistent with slow penetration of the adsorbed P.