The aim of this work was to determine the role of saliva in wine aroma release by using static and dynamic headspace conditions. In the latter conditions, two different sampling points (t = 0 and t = 10 min) corresponding with oral (25.5 °C) and postoral phases (36 °C) were monitored. Both methodologies were applied to reconstituted dearomatized white and red wines with different nonvolatile wine matrix compositions and a synthetic wine (without matrix effect). All of the wines had the same ethanol concentration and were spiked with a mixture of 45 aroma compounds covering a wide range of physicochemical characteristics at typical wine concentrations. Two types of saliva (human and artificial) or control samples (water) were added to the wines. The adequacy of the two headspace methodologies for the purposes of the study (repeatability, linear ranges, determination coefficients, etc.) was previously determined. After application of different chemometric analysis (ANOVA, LSD, PCA), results showed a significant effect of saliva on aroma release dependent on saliva type (differences between artificial and human) and on wine matrix using static headspace conditions. Red wines were more affected than white and synthetic wines by saliva, specifically human saliva, which provoked a reduction in aroma release for most of the assayed aroma compounds independent of their chemical structure. The application of dynamic headspace conditions using a saliva bioreactor at the two different sampling points (t = 0 and t = 10 min) showed a lesser but significant effect of saliva than matrix composition and a high influence of temperature (oral and postoral phases) on aroma release.