Phosphorus fertilization can increase P losses in surface runoff, but limited information is available for fields with <2% slopes in the US Midwest. Our objectives were to determine the effects of tillage-fertilizer placement (no-till-broadcast, strip till-broadcast; or strip till-deep placement, -15-cm subsurface band) and fertilizer rate applied in the fall (0, 52, or 90 kg P2O5 ha-1 yr-1) on runoff P concentrations and loads in fields with <2% slopes near Pesotum, IL, during fall and spring simulation runoff events, and to measure corn (Zea mays L.) and soybean [Glycine max (L.) Merr.] grain yield. Across four simulated runoff events, deep placement reduced dissolved reactive P (DRP) loads by 69 to 72% compared with the broadcast treatments. A tillage-fertilizer placement × P rate interaction showed that DRP and total P (TP) concentrations remained low when P was deep placed, regardless of P rate, whereas concentrations increased with increasing P rate for the broadcast treatments, but no differences existed for bioavailable P (BAP) (α = 0.05). At one site, rainfall simulation in the spring versus fall increased runoff volumes but sharply decreased BAP concentrations. During fall runoff simulations, deep placement reduced TP loads, and greater TP loads occurred with the 90- than the 52-kg P2O5 ha-1 yr-1 rate. Similarly, when P was broadcast in the fall, DRP and TP concentrations were greater than deep-placed P, but no treatment differences occurred in the spring. Deep banding P and K did not reduce crop yield but reduced runoff losses of P from flat fields compared with broadcast P applications, particularly at high rates of P application.