The present work aimed to estimate the probability of bottled white wine exhibiting acceptable colour change during a certain predefined time using a predictive model based on a descriptive mathematically expressed theory. Nine mono-varietal Greek white wines (produced by three varieties) were bottled and stored at two different temperatures (20 and 30 °C). Their absorbance at 420 nm (A420) was recorded every three months for 12 months. Results were processed by applying the root cause analysis versus storage conditions. Furthermore, a predictive mathematical model was adapted to the experimental data based on the probability of the product showing browning-index values higher than a specific threshold during its shelf life. As expected, the absorbance at 420 nm increased with time, while the elevated storage temperature had a significant effect on colour alteration. The browning rate was dependent on grape variety. When storage temperature was 20 °C, a lower A420 increase was observed for Sauvignon blanc wine, while at 30 °C, browning was higher. Assyrtiko wines were more resistant to colour change at both temperatures. The probability of a bottled product maintaining an acceptable colour change by the end of one year, was used as a measure of its quality. A very successful fit was achieved between mathematical and experimentally obtained data for shelf life predictions. The modelling of the experimental results, as well as the exploitation of this fitting to make predictions independently of the specific results, indicate that storage temperature will potentially impact the colour of wines over time, depending on the evolution of the absorbance values at 420 nm. For the wine sector, the selection of the appropriate storage conditions for a given grape variety is an important factor to take into consideration.