Herein, the phenolic composition and colour attributes of red grapes extracts (obtained with a fast methodology) were correlated with those of their corresponding wines to predict the final quality properties of wines. The phenolic parameters were evaluated as total phenolic compounds (TPC), total anthocyanins (TA) and total condensed tannins (TCT), whereas the chromatic parameters were evaluated as colour intensity (CI), tonality (To), and the percentages of yellow, red and blue tones. All of them were determined by usual UV–Vis spectrophotometric methods. To get robust models, grapes of five red varieties were collected at three different ripening stages throughout the 2009 vintage. Good correlations between the results from grapes and wines were obtained, showing high regression coefficients and low prediction errors for TPC (R2 = 0.929, RMSE = 5.99%), TA (R2 = 0.953, RMSE = 7.23%) and CI (R2 = 0.954, RMSE = 7.58), concluding that these wine phenolic properties can be predicted reliably from the extracts obtained with an optimised fast extraction method from grapes on the ripening controls along the maturity process.