Abstarct: Nowadays, Tunnel Boring Machine (TBM) is widely used in tunneling projects. Estimation of TBM performance is one of the most important parameters in estimation of cost and time of tunneling projects. TBM performance is evaluated using Advance Rate (AR), Penetration Rate (PR) and Utilization factor (U). Various models and method for estimation of these parameters were presented The overall purpose of this thesis is to use probabilistic methods for estimation of TBM advance rate. For this purpose, the equations of QTBM, NTNU, CSM, Hassanpour (2011) and Farrokh (2012) models as performance function in three methods Monte Carlo Simulation (MCS), Point Estimate Method (PEM) and Taylor series are used. At First, according to data of lar water conveyance tunnel lot-2, for input parameters of models, proper distribution function was considered, then with the implementation of the Monte Carlo Simulation, the advance rate was obtained with 95% certainty. In PEM and Taylor series methods, equations of these methods were written in MATLAB, and then with the implementing of code, by entering the mean and standard deviation of the parameters and equations of each model, the mean and standard deviation of the advance rate was calculated. Among models, for the geological units including tuffy shale and marl rocks, the average of the advance rate calculated by the modified CSM model with EMI correction factor and in the geological units including gabbro, sandstone, siltstone rocks, and crushed zone, the results of Farrokh model was closer to the average value of the actual advance rate. In all the geological units, the NTNU model has a much lower advance rate than other models. Among the probabilistic methods, the average of the advance rate calculated by the MCS method was closer to the average of the actual advance rate. Also, the advance rate calculated by deterministic analysis (with using most-likely value of parameters in different models). The average and value of the advance rate calculated by the MCS method and deterministic analysis, except in the QTBM model, which was calculated with two different methods, there was not much difference in the other models. In the deterministic analysis, only a mean value for the advance rate was calculated, whereas, in probabilistic methods, the calculated advance rate was histogram with a confidence level. Using probabilistic methods in combination with TBM prediction models, with considering different states, helps to estimate the range of advance rates more confidently.