Previous efforts at turfgrass model calibration have raised questions about the applicability of using small plot research data to evaluate model performance. In 2003 a runoff research facility was constructed at the University of Maryland to examine the relationship between plot size and agrochemical runoff. The facility consists of three small (3.6 m x 9.1 m) and large scale (12.2 m x 38.1 m) creeping bentgrass (Agrostis stolonifera) plots located on a 3.5% hillside slope. Hydrographs were generated in 2005 and 2006 by simulating a 36 mm hr^-1rainstorm for the time needed to initiate runoff plus 90 additional minutes. Granular forms of urea and treble superphosphate as well as a tank mix of Three-Way Bentgrass Selective (2,4-D, MCPP, dicamba), Prostar (flutolanil) and Dursban (chlorpyrifos) were applied to the plots one day before the simulated rainstorm. Water samples were collected every five minutes to assess nutrient and pesticide runoff losses. Results obtained from the simulated rainstorms indicate that turf plot size has little effect on unit area runoff losses of total nitrogen, 2,4-D, flutolanil and chlorpyrifos. Unit area runoff losses of total phosphorus however are greater from large size plots than from small size plots. Total suspended sediment data collected during the runoff events suggest that the different phosphorus loss rates seen in the two size plots is not due to different amounts of sediment being lost from the two size plots. We believe the divergent results obtained for total P losses can be attributed to more extensive overland stream-like flow seen in the large plots coupled with the low water solubility of treble superphosphate granules.