A general probabilistic fatigue crack growth prediction methodology under random variable loading is developed using a novel equivalent stress level model and the inverse first-order-reliability method (IFORM). The proposed equivalent stress level model is based on the equivalent transformation of a random variable loading to constant amplitude loading, which avoid cycle-by-cycle calculation. An inverse first-order reliability method (IFORM) is used to evaluate the fatigue crack growth at any arbitrary reliability level. Inverse FORM method reduces the number of function evaluations and the computational cost is significantly reduced. The proposed method is very suitable for real-time damage prognosis and on-line decision making. Numerical examples are used to demonstrate the proposed method. Various experimental data under variable amplitude loading are collected and model predictions are compared with experimental data for model validation.