Earthquake is one of the main hazards of natural environment which threatens the dams' safety. Strong ground motion parameters are studied and the results are used in the seismic design of dams to investigate probabilistic seismic hazard analysis. These results are the basis of codes and seismic design handbook of important structures such as dams. Seismic hazard analyses can be performed utilizing deterministic approach as well as probabilistic one. In the deterministic approach, peak ground motion is determined which is used as single parameter for later designing. However, in the probabilistic approach, which was used in the current research, a distribution for the annual rate of exceedance at different ground motion levels is presented. Generally, the parameters used in the seismic hazard analysis are uncertain which results in uncertain outputs. Many different methods are developed to consider the input data uncertainty. Since, there is an increasing demand for more accurate criteria which results in acceptable decision making, researches have sought different sources of uncertainty and attempted to quantify them. Uncertainty can be divided into two main grou aleatory and epistemic uncertainties. Some parameters in the seismic hazard analysis like the time of occurrence and its location have aleatory uncertainty while some others like attenuation models or seismic source models have epistemic uncertainty. Current research, investigates the uncertainty of seismic source location as an epistemic uncertainty phenomenon by using Monte Carlo simulation. The aim of this study is to explain the procedure of quantifying uncertainty of seismic source location and the effect of this uncertainty on the seismic hazard analysis is then evaluated. In this regard, a problem of a riverbed as the potential location of dam construction was investigated. The fault locations around the riverbed were presumed to be uncertain. This uncertainty was modeled in two different modes including rotation and translation of the faults segments. By assuming a Normal distribution for the dispersion of fault location, probabilistic seismic hazard analysis accompanied by Monte Carlo simulation was performed. Studying the results distribution along the stream line, the locations with less uncertainty were identified. Furthermore, the results of different modes (translation and rotation) were compared. As expected, it was observed that translation mode of the faults results in more dispersed results. In order to examine the response of the riverbed section due to propagating SH waves, a semicircular valley is considered. Monte Carlo simulation was utilized to model the incident wave dispersion and determination of critical points of the valley. In another example, the combination of probabilistic seismic hazard analysis and Monte Carlo simulation was performed for a case study (Behjat Abad dam site). The coefficient of variation and standard deviation of the resulted distribution were computed for different annual rate of exceedance and return periods. Based on these results, design level of DBL (Design Based Level) and MDL (Maximum Design level) were proposed. These parameters depend on desirable reliability level. The results of two modes; translation and rotation, for surrounding faults of this dam site, were compared to each other. Finally, it was concluded that the sensitivity of the results directly depend on the input parameters, their distribution (its type and the degrees of dispersion) and the amount of sampling points in the Monte Carlo simulation. Keywords probabilistic seismic hazard analysis, uncertainty, seismic source dislocation, Monte Carlo.