Phosphorus enrichment of surface water is a concern in many urban watersheds. A 3-yr study on a silt loam soil with 5% slope and high soil test P (27 mg kg(-1) Bray P1) was conducted to evaluate P fertilization and clipping management effects on P runoff from turfgrass (Poa pratensis L.) under frozen and nonfrozen conditions. Four fertilizer treatments were compared: (i) no fertilizer, (ii) nitrogen (N)+potassium (K)+0xP, (iii) N+K+1xP, and (iv) N+K+3xP. Phosphorus rates were 21.3 and 63.9 kg ha(-1) yr(-1) the first year and 7.1 and 21.3 kg ha(-1) yr(-1) the following 2 yr. Each fertilizer treatment was evaluated with clippings removed or clippings recycled back to the turf. In the first year, P runoff increased with increasing P rate and P losses were greater in runoff from frozen than nonfrozen soil. In year 2, total P runoff from the no fertilizer treatment was greater than from treatments receiving fertilizer. This was because reduced turf quality resulted in greater runoff depth from the no fertilizer treatment. In year 3, total P runoff from frozen soil and cumulative total P runoff increased with increasing P rate. Clipping management was not an important factor in any year, indicating that returning clippings does not significantly increase P runoff from turf. In the presence of N and K, P fertilization did not improve turf growth or quality in any year. Phosphorus runoff can be reduced by not applying P to high testing soils and avoiding fall applications when P is needed.