Recent changes in soil testing methodology, the important role of P fertilization in early establishment and soil coverage, and new restrictions on P applications to turf suggest a need  for soil test calibration research on Kentucky bluegrass (Poa pratensis L.), tall fescue (Festuca arundinacea Schreb), and perennial ryegrass (Lolium perenne L.).  Greenhouse studies were conducted to examine the relationship between soil test P levels and P needs for grass establishment using 23 New Jersey soils with a Mehlich-3 extractable P ranging from 6 to 1238 mg kg^-1.  Mehlich-3 extractable P and Al were measured by ICP to evaluate the ratio of P to Al as a predictor of need for P fertilizer.  Kentucky bluegrass establishment was more sensitive to low soil P availability than tall fescue or perennial ryegrass.  Mehlich-3 was an effective predictor of need for P fertilization but the ratio of P to Al (Mehlich-3 P/Al %) was a better predictor of tall fescue and perennial ryegrass establishment response to P fertilization than soil test P alone.  The Mehlich-3 soil test P critical level for clipping yield response was 280 mg kg^-1 for tall fescue and 231 for perennial ryegrass.  The Mehlich-3 P/Al (%) critical level was 42% for tall fescue and 33% for perennial ryegrass.  Soil test critical levels, based on estimates from clipping yield data, could not be determined for Kentucky bluegrass using the soils in this study.  Since different grass species have varying critical P levels for establishment, both soil test P and the species should be incorporated into the decision-making process regarding P fertilization.