One of the essential aspects of beer is its characteristic taste which is not stable and varies with time and storage temperature, due to the changes in its composition. The quality of beer can be measured through sensory evaluation by a trained panel, but the quality assesment could also be accomplished by evaluation of the physical chemistry components. The objective of this study was to predict the concentration of the physical chemistry components and the sensory quality of the bottled beer, while being stored at different temperatures and time intervals. For this purpose, beer was produced in a local industry with two concentrations of oxygen (0.02 and 0,15 p.p.m). For each of them 384 samples were bottled and stored at 5°C, 28°C and 60°C. Initially and at intervals of seven days over a sixteen week period, the content of carbon dioxide, pH, color, turbidity, TBA, and bitterness were measured by instruments. Simultaneously, eight trained panelits did a sensory evaluation of grade of oxidation, quality of bitterness and general acceptance. The data obtained was divided in two groups, one was used to find the prediction models and the other was used to validate these models. A multiple linear regression was used to estimate the coefficients of the models. Theses models explained 86÷88 percentage (p < 0.01) of the changes in the compositions and sensory characteristics. The general expression was Y equal B(0) + B(i) (X)i where Y is the concentration of components or the sensory characteristis and X(i) is the oxygen concentration, the temperature and the storage time. The prediction models were validated by an algorithm written in Matlab language. The results suggest that the concentration of components and the sensory characteristics of bottled beer could be predicted by oxygen concentration and storage conditions.