The paper shows a procedure to explore and optimize the relationship between raw material characteristics (fruit ripening and storage grade) and a multidimensional index representing virgin olive oil quality. The utilized procedure has required five different steps: a) flavour profile evaluation of several commercial olive oil samples; b) perceived quality evaluation by COI method; c) selection of sensory variables related to perceived quality; d) identification of chemical markers related to selected sensory variables, e) combination of chemical markers into an olive oil quality index (overall desirability function-D). Utilizing the Central Composite Design, nine oil extractions were made in order to obtain a model able to relate oil quality to olive fruit characteristics (ripening and storage). The procedure proposed, in attaining quality index (desirability function-D), allows to optimize the extraction yield (amount of oil), the oil stability (amount of phenols), and the consumer oil perception related to the amount of trans-2-hexenal and phenols.