Dimensionless experimental formulae based on a rational reciprocal function for correlation of flashpoint data of binary mixtures of two flammable components have been developed. The formulae are based on data obtained from flash-point experiments. The proposed approach requires only two coefficients, molar fraction of components and flashpoint temperatures of the pure flammable components to be known in advance. Literature data were used for formulae verification and validation obtained results indicate that accuracy is comparable and to some extent better than that of conventional flash point prediction models. Dimensional analysis and scaling of data have been performed in order to define the correct construction of the equation fitting flash-point data in dimensionless form using the independent variables suggested by Catoire. Stefan number relevant flash-point of a single compound or a blend has been defined.