The overall quality of virgin olive oil (VOO) is closely related to its oxidative stability that is usually evaluated through the stability index measured by the Rancimat apparatus. Quality characteristics and also pro-oxidant and antioxidant content for 52 Greek VOO samples (Koroneiki cv) were used to build up a model capable of predicting stability. Collinearity diagnostics, variable selection, and regression analysis were applied to the experimental data to locate the contribution of each parameter to the keeping quality of the samples. The predictive ability of the model was confirmed for a second VOO ample set of the same cultivar. It was found that except for the peroxide value, which negatively influences the stability, other important parameters were α-tocopherol, total polar phenol and total chlorophyll content. It is concluded that the colorimetric determination of total polar phenols, the spectrometric determination of total chlorophylls and the high-performance liquid chromatography analysis of α-tocopherol, not presently included in the established methods of official analysis, can be used for a better evaluation of VOO quality. These parameters, which can be easily adopted as routine methods by the industry, seem to be of utmost importance for shelf life prediction and expiration dating if applied for the promotion of the most competitive products in the international olive oil market.