The pasteurization is employed for extending shelf-life and keeping the quality of products constant. However the pasteurization undesirably accelerates the oxidation of beer which renders volatile compounds concerning off-flavor. The pasteurization conditions was optimized to avoid off-flavor of beer during pasteurization. Under the isothermal condition, thermal destruction kinetics such as D and Z value of Lactobacillus brevis which is reported the most common beer spoilage, was determined. The binomial data (detected or non-detected of off-flavor) was treated with logistic regression to estimate off-flavor development (OFD) times. The temperature dependence of OFD times was established in terms of Arrhenius relationships. Optimized pasteurization temperature and time were found at which OFD times was not detected. The constraint for optimization was that the pasteurization degree should be larger than 5 decimal reduction time. The optimization was conducted through mathematical simulation using kinetics and temperature-dependent models for microbial death and OFD times. The optimized results were validated by the corresponding experimentations, which met the requirements that the concentration of Lactobacillus brevis was 5-Log reduction and the OFD times not detected.