A probability-based procedure has been developed to predict reliability of gas turbine engine blades subjected to high cycle fatigue. The procedures provides a systemic approach for predicting and designing turbomachinery blading reliability against various potential fatigue problems for all relevant vibratory modes and taking into account variability in geometry (e.g. dimensional variation, mistuning, etc.), unsteady aerodynamics, structural damping, and thermal loading. The variability in materials (e.g. damage, cracks, degradation, etc.) can be also considered in this approach. A reliability prediction was performed on gas turbine blades at high frequency modes (e.g. third strip modes) using a probabilistic vibratory stress distribution in conjunction with the Haigh or modified Goodman Diagrams. The cumulative reliability and risk assessment are then calculated using the fast probability integration (FPI) technique to construct a novel probabilistic Goodman (or Haigh) diagram which provides blade’s lifetime design guide line and an optimal maintenance strategy in management in decisionmaking relating to the PM/inspection scheduling, replacement, spare parts requirements.