The study of the elemental and carbon isotopic composition of dissolved organic matter (DOM) is of great interest in deciphering the origin and transformations of organic matter in estuarine and coastal environments. Styrenedivinylbenzene copolymer (PPL) sorbent is currently the most commonly used material for the isolation of DOM from environmental samples. It is notably used for the development of molecular formula-based indices used to study DOM reactivity. However, any extraction procedure (including with PPL) will fractionate the DOM. If this fractionation is not well constrained it can lead to biased interpretations of the biogeochemical processes affecting DOM. In this work we investigate the fractionation effects of the PPL sorbent on the size class distribution of DOM and the carbon isotopic composition of the PPL retentate. The use of size exclusion chromatography, that does not require a pre-concentration step, allows a precise study of the fractionation of DOM (including aromaticity) by the PPL resin. Extractions performed on two types of humic substances dissolved in artificial seawater, using the PPL resin, showed high extraction yield (> 85%) and the stable isotopic carbon composition (δ 13 C) of these compounds was successfully recovered. These results indicate that salinity is not a parameter affecting extraction yield on PPL sorbent. For a hydrophilic compound (atropine) the extraction efficiency was low (33%) and δ 13 C signature was underestimated. Size exclusion chromatography measurements, in samples collected along a salinity gradient, demonstrate that the PPL sorbent strongly fractionates DOM. Although the DOM size class distributions in freshwaters and marine waters were initially different, their retentates were marked by similar size class distributions. This work demonstrates that PPL resin captures DOM compounds with less nitrogenous content and it seems to have a lower affinity for aromatic compounds of marine origin than of terrigenous origin. The study of DOM distribution in a macro-tidal estuary demonstrated the limitations of PPL resin extraction in capturing an internal N-rich DOM production event at the time of sampling. Futhermore, the isotopic composition of the PPL resin retentate appears to depend on the extraction efficiency of the more hydrophobic compounds which changed along the salinity gradient. This study recommends careful interpretations of data that only rely on PPL extractions, particularly for works tracking the origin of DOM in estuaries and comparing DOM composition across ocean biogeochemical domains.