Soil test correlation is continuously necessary to affirm critical soil test values (CSTV) for specific nutrients, as changes in crop management and development of new plant varieties may change CSTVs. The objective of this study was to determine the CSTVs of phosphorus (P) and potassium (K) for soybean [Glycine max (L.)Merr.] and corn (Zea mays L.) using three long-term trials in North Carolina. Soybean was cultivated in 2020 and corn in 2021 in three long-term trials established on research stations located in the Tidewater, Coastal Plain, and Piedmont regions of NC. In each trial, up to five rates of P (0-88.2 kg P ha −1 ) and K (0-186.8 kg K ha −1 ) were applied annually at planting. Soil and tissue samples were analyzed and yield measured. There was yield response to P in all site-years. The average CSTVs of P for corn and soybean were 51, 66, and 14 mg kg −1 for the sites at Tidewater, Coastal Plain, and Piedmont regions. The CSTV of P for Piedmont (14 mg kg −1 ) differs greatly from the current recommendation for NC (52 mg kg −1 ). The response to K fertilization was observed only in three site-years and the CSTVs varied from 49 to 93 mg kg −1 , while the current recommendation is setting the CSTV of K at 85 mg kg −1 . These results indicate it is necessary to develop further studies of soil test correlation for P and K in North Carolina to better estimate the CSTVs for the state.