To safely handle, transport, and store flammable or combustible liquids, such as biodiesel and its blends, it is important to have knowledge of a few physical-chemical properties. The Flash Point is an important one, as it is related to the flammability of the fuel blend. It can be experimentally measured through open cup or closed cup standard procedures. However, due to the usual scarcity of experimental data for multicomponent systems, developing a model to predict flash points of mixtures is of interest. To do so, there are a few possible approaches, which include empirical regression of data, vapor pressure-based methods, and QSPR. When it comes to mixtures, the most popular modeling method is based on vapor pressure, which usually employs LeChatelier's rule and vapor−liquid equilibria (VLE) calculations to flash point prediction. Generally, a γ−φ approach is adopted to describe the VLE behavior, although some authors have shown interest in φ−φ approaches. In recent years, studies on QSPR for FP prediction of mixtures have evolved, which represents an advance toward more generalized FP prediction models. Additionally, COSMO type models have been gaining attention in FP prediction, usually associated with vapor pressure models or even empirical models. When it comes to biodiesel though, not much progress has been made since 2014, with just a few works being published since then. This paper seeks to review advances made in FP prediction methods for mixtures in general, while giving attention to those involving biodiesel and petro-diesel.