Storage is important for virgin olive oil, a product obtained only during the harvest period, which requires a year-round storage until its best-before date. Low temperatures slow undesirable reactions, though this method is not widely applied. The objective of this paper is to assess the impact of frozen storage on the volatile composition and sensory properties of virgin olive oils. The quality parameters, volatile composition, and sensory profiles were analyzed for samples stored under different conditions (time 0, sixth month supermarket and frozen storage, and long-term-frozen). The physicochemical parameters of the samples stored under supermarket conditions showed significant differences (p < 0.05), with the frozen-storage sample after three months of storage. Additionally, the samples stored under supermarket conditions showed higher volatile concentrations than frozen ones, with increased concentrations of aldehydes and acids producing sensory defects. Thirty-two samples, considered as the long-term-frozen, were divided into three groups depending on the frozen-storage time (1, 6 or 10 years). These long-term-frozen storage samples confirmed the suitability of the proposed oxidation markers (pentanal, hexanal, heptanal, nonanal, acetic, propanoic, butanoic, and hexanoic acids) for differentiating storage conditions and times. This work highlights the oxidation process under different storage conditions and suggests oxidation markers.

The quantification of volatile compounds in virgin olive oil poses several analytical challenges due to the existence of different concentrations, chemical families, and the possible matrix effect. Accurate quantification, using adequate methodological calibration and statistical procedures, is essential for obtaining reliable results. The aim of this work was to develop and validate an analytical–statistical approach for the quantification of volatile compounds in virgin olive oil. Therefore, several analytical parameters were determined for four calibrations. The ordinary least square (OLS) linear adjustment was selected over the weighted least square due to the homoscedasticity of the variable errors. Additionally, standard addition (AC) and AC with an internal standard (IS) exhibited greater variability, whereas external matrix-matched calibration (EC) was identified as the most reliable approach for quantifying volatile compounds in virgin olive oil. The employment of an IS did not improve the performance of the method in any case. Thus, based on the statistical results, the OLS linear adjustment with EC was selected as the best statistical–analytical approach for quantifying volatiles in olive oil matrices. The volatiles of nine virgin olive oil samples were quantified, applying different methodological calibrations, and no differences were detected, underscoring EC as a superior alternative.