Abstarct: Tunnel Boring Machine (TBM) performance prediction is one of the most important procedures to evaluate the time and cost of tunneling projects. In this research, TBM penetration rate was predicted by Deterministic and Probabilistic Rock Engineering System (RES) methods baxsed on a databaxse developed along the 8.7 KM long Karaj- Tehran water conveyance tunnel. In this respect , nine input parameters belonging to the geology condition and five parameters belonging to the TBM were selected and included in the models. Geological parameters consist of UCS, BTS, number of joint per volume, the angle between discontinuities and tunnel axis weathering, brittle index, ground water condition, mixed face and quartz content and TBM parameters consist of cutterhead speed (RPM), the spacing of cutters, thrust force, torque and power. The results obtained from Deterministic and RES models to predicte the penetration rate and its comparison with measured penetration rate in the field showed a good compability in some sections along the tunnel. The correlation coefficients were 0/4885 and 0/4554 for Deterministic and RES models, respectively. The overall outcome showed that the TBM performance is a nonlinear and complex problem.