China is one of the most earthquake-prone countries in the world and research on earthquake engineering has long been highly valued by various scholars. Ground motion attenuation relationship is a significant and useful topic in engineering seismology. It provides the background for aseismic design of structures and evaluating the earthquake hazards. Therefore, it is very significant to obtain a suitable ground motion attenuation equation for high seismic areas. Bayesian method is being widely used for engineering fields since Bayesian probabilistic approach provides a rigorous solution to parametric identification and uncertainty quantification by using probability logic and can propose a plausible model class which takes into account the robustness and data fitting capability simultaneously. It has been proved that it is efficient approach to deal with regression analysis and to select the most plausible model class among a set of model class candidates based on measured data. The Boore-Joyner-Fumal formula which estimates the peak ground acceleration is one the most well-known models. In this thesis, the complexity for this formula is investigated and efficiency-robustness balanced formula is proposed by utilizing Bayesian approach. A database of strong ground motion records from Wenchuan Earthquake is used for the analysis.