Many state and local governments have enacted regulations to prevent phosphorus (P) application to turfgrass unless an application is required by a soil test.  The Mehlich-3 soil extraction with inductively coupled plasma optical emission spectrometry (ICP-EOS) of P is extensively used to determine available soil P, however calibration studies for established sand greens are lacking.  The purpose of this study was to evaluate various agronomic responses of turfgrass to different soil test P (STP) levels. In 2008, a creeping bentgrass (Agrostis stolonifera) putting green was constructed at the O.J. Noer Turfgrass Research Facility in Madison, WI using 30 cm of non-amended calcareous sand overlaying pea gravel.  Thirty-two plots were treated with various amounts of monopotassium phosphate to create a broad range of STP values which initially ranged from 4 to 55 mg P kg^-1 soil.  One set of 16 plots was treated with trinexapac-ethyl while the other served as 16 controls.  Clipping yield, Mehlich-3 STP level, tissue P content, turfgrass visual quality rating, and color index were quantified monthly during the 2009 and 2010 growing seasons.  Nonlinear regression analysis was used to calculate both STP and tissue P content critical levels.  The STP level of all plots declined during the 18 month study period with largest decreases occurring in plots with high initial STP levels.  Turfgrass quality was sensitive to both STP and tissue P content and provided high resolution critical values with linear plateau regression.  The Mehlich-3 STP and tissue P content critical levels for turfgrass quality ranged from 6 to 11 mg kg^-1 and 1.5 to 4.4 g kg^-1 during the growing season, respectively.  Trinexapac-ethyl inconsistently altered the STP critical level and never by practical level.  These Mehlich-3 critical values are below values used by soil testing laboratories and interpretations published in popular press articles and textbooks.