: In this thesis, PBI-based organic-inorganic nanocomposite membranes (PBI-Fe 2 TiO 5 (PFT), PBI-La 2 Ti 2 O 7 (PLT) and PBI-SrCeO 3 (PSC)) were prepared with different mass percentages of the nanoparticles and then the nanocomposite membranes were doped with phosphoric acid in order to show the property of proton conductivity. The structure of the resulting nanocomposite membranes were examined by SEM-EDX, XRD and FT-IR analyses and the results showed the homogenous dispersion of the inorganic nanoparticles within the PBI polymer matrix. Furthermore, investigation of the mechanical and chemical stability and phosphoric acid leaching of the nanocomposite membranes showed the applicability of PFT, PLT, and PSC nanocomposite membranes in high temperature PEM fuel cells. The highest phosphoric acid doping levels of the nanocomposite membranes were found for PFT 4 , PLT 8 and PSC 8 membranes that showed the maximum proton conductivities of 0.078, 0.082, and 0.105 S/cm, respectively. These membranes were used in a single cell of membrane-electrode assembly (MEA) for application in high temperature PEMFCs. The highest power and current density were found for PSC 8 , which showed the highest proton conductivity among the nanocomposite membranes. In addition, proton conductivity of the phosphoric acid doped nanocomposite membranes was predicted using three models (non-Arrhenius model, Bruggeman model, and Choi model). The results showed that the proton conductivities predicted by the non-Arrhenius model had a good agreement with the experimental data, but this model does not consider the effects of structural properties on the proton conductivity of the nanocomposite membranes. The effects of phosphoric acid doping level (PA dop ), volume fraction of phosphoric acid in the membranes (? PA ), mass fraction of the nanoparticles (w d ) and temperature on the proton conductivities predicted by the semi-empirical Bruggeman model has been considered. The proton conductivities predicted by this model for PFT and PLT membranes showed relatively low relative errors (2.70 and 4.23% , respectively), while a higher relative error (7.21%) was obtained for PSC membrane. Therefore, the developed Bruggeman model can be used for a relatively good estimation of the proton conductivity of the nanocomposite membranes PFT, PLT and PSC. In adition, the effects of structural characteristics of the membranes are well considered in the Choi model which have more complexibility rather than the other models. In this model, porosity or volume fraction of phosphoric acid inside the membranes structure, proton concentration, and tortusity factor of the membranes were determined theoretically and proton diffusion coefficient was predicted theoretically-experimentally (with the aid of Gaussian software). The proton conductivities predicted by this model for the nanocomposite membranes PFT, PLT and PSC showed relatively lower errors (1.99% for PFT, 4.33% for PLT, and 2.68% for PSC) than those of Bruggeman model. So, because the proton conductivities obtained by this model had lower relative errors and this model condiders the structural properties of nanocomposite membranes as well as the proton hopping mechanism for proton transfer along the membranes, it has a better capability for prediction of proton conductivity of the nanocomposite membranes. Keywords: Organic-inorganic nanocomposite membranes, PBI, Proton conductivity, High temperature proton exchange membrane fuel cell (HT-PEMFC), Modeling.