Grape phenolics represent a class of compounds of primary interest for winemaking, as they are responsible for main sensory characteristics and important physiochemical interactions with other chemicals. These compounds are extracted from the solid phase of the berry, such as skin and seeds, throughout the alcoholic fermentation and can be adsorbed by insoluble solids, such as plant cell walls (CW) and on the surface of seeds and yeast cells. The extent and rate of these molecules’ extraction strongly depend on temperature, alcohol concentration, and juice-mixing procedures. CW composition of grape skin has also been shown to play a key role in determining the extractability of grape phenolics.In this study we investigated the relationship between grape skin cell wall composition and the extraction of grape phenolic compounds. While doing so, we evaluated the opportunity of applying spectral methods to quickly and non-destructively predict cell wall composition and/or extractability of phenolics. Grapes were collected and micro-fermentations carried out at standardized conditions. Grape phenolic composition was analyzed by exhaustive extraction and high-performance liquid chromatography (HPLC). Wine phenolic profile was characterized by HPLC. CW material was isolated from berry skins as alcohol-insoluble residue and characterized. Protein, phenolics and pectin (analyzed as uronic acid) were determined applying colorimetric procedures. Klason lignin was measured gravimetrically. Intact berries were then scanned with Ocean Insight spectrometers, working on a wavelength bandwidth range from 350 to 2,200 nm. Grape phenolics and their extractabilities were found to mainly depend on the variety, the site and its climate. The relative amount of isolated cell wall material was found to follow a variety-dependent trend. A similar effect was found for CW proteins and phenolics, whose content negatively impacted anthocyanin and phenolic extractability. Isolated cell wall, CW lignin and pectin did not seem to correlate with phenolic extractability. With the data analysis carried out, wavelengths in the range 1004.7-1007.0 nm were found to be successful candidates to predict polymeric pigments content and extractability in wine.