<p>Spin-trapping agents-based electron paramagnetic resonance (EPR) is still widely used to detect hydroxyl radicals (•OH) in engineered environmental systems. Conventionally, the “four-line peak” of DMPO/•OH (1:2:2:1) was considered the gold standard for the presence of •OH, and signal intensity was occasionally applied to quantify •OH concentrations. Based on chemical reaction networks and reaction rate constants, we established a network dynamics model to quantitatively determine the concentrations of •OH and SO₄^•−. For example, when persulfate (S₂O₈²⁻, 100 mmol/L, final pH = 3.33) was activated by FeS₂ (100 g/L), SO₄^•− concentration was 2.66 × 10⁻¹⁰ mol/L, 6 orders of magnitude higher than that of •OH (2.67 × 10⁻¹⁶ mol/L), while the concentration of DMPO/SO₄^•− (3.14 × 10⁻¹¹ mol/L) was 7 orders of magnitude lower than that of DMPO/•OH (2.34 × 10⁻⁴ mol/L). These results were validated by EPR. Our study revealed that 81.1%–81.5% of DMPO/•OH is derived from DMPO/SO₄^•− hydrolysis and only 18.5%–18.9% is from direct capture of •OH, questioning the reliability of detecting •OH based on the appearance of the “four-line peak”. Our study underscores the necessity of considering all the transformations among radicals and their adducts during EPR analysis, which also provides a direct and effective method for detecting other radicals with extremely short half-lives in other heterogeneous persulfate systems. The high sensitivity of SO₄^•− and •OH to pH also provides an avenue to regulate the generation of reactive species.</p>