With the increase of the demand of low flavouring smoked seafood products, there is a need of methodologies able to distinguish between different seafood treatments, as not all of them are allowed in all markers. Following this objective, in the present work an untargeted volatolomics approach was applied to identify volatile markers that demonstrate that Cold smoked products can be distinguished from Tasteless smoke neither Carbon monoxide treated seafood, which are prohibited in the European Union. The use of dynamic headspace for the volatile extraction followed by thermal desorption in combination with Gas Chromatography (GC) coupled to single quadrupole Mass Spectrometry (MS) has been employed for the determination of volatile composition of smoked fish. Data processing consisted on the use of PARADISe software, applied for GC/MS data treatment, followed by the multivariate analysis with PLS_Toolbox (MATLAB), and finally the creation and validation of statistical classification model. All 107 variables obtained allowed the construction of a model reaching the correct classification of 97% of the blind samples, while a simplified model with only 11 variables correctly classified up to 93% of the blind samples. These 11 compounds were elucidated to develop subsequent target volatolomics approaches, if needed. Ordered according to the importance in the classification model, the elucidated compounds were: 3-methylcyclopentanone, ethylbenzene, 2-methyl-2-cyclopenten-1-one, 2-methyl-benzofuran, furfuryl alcohol, 2-acetylfuran, acetophenone, guaiacol, 1-hydroxy-2-butanone, 4-vinylguaicol and acetoin. The results demonstrated the great potential of untargeted volatolomics for smoked seafood treatments classification.