<p>The difference between the phosphorus (P) levels determined by inductively coupled plasma optical emission spectroscopy (ICP-OES) and colorimetry in soil extracts is termed molybdate-unreactive P (MU-P). The dynamics and characterization of MU-P in natural environments remain not fully elucidated, making it essential to understand how different extractants for available P influence MU-P levels and which soil attributes and P forms are associated with this pool. This study compared three extractants used in soil P tests for their ability to access MU-P, explored associations between soil chemical attributes and MU-P, and evaluated whether there is a relationship between P species obtained by XANES and MU-P in soils of contrasting nature. To achieve this, 28 representative soils were collected in non-cultivated forestry areas (native or reforested), which had never received anthropogenic P inputs. On average, Mehlich-3 extracted the highest amounts of MU-P; however, the correlations between Mehlich-3 MU-P with other chemical attributes were weaker than those observed for Mehlich-1 and resin. XANES analysis did not reveal a clear relationship between MU-P levels and P-species in the soil samples. Among soil attributes, organic carbon (OC) displayed the strongest correlations with MU-P, which highlights its potential as a predictive attribute for MU-P in natural environments. Managing OC in soils could influence MU-P levels, affecting phosphorus availability for plants. To this end, it is important to consider the P extractant, with Mehlich-1 and resin being better options in relation to Mehlich-3. There are limitations to XANES K-edge in effectively capturing the complexity of MU-P in soil samples.</p>