This paper introduces a maintenance decision-making strategy in the general area of replacement and reliability of mechanical components. The decision-making strategy involves the optimization of replacement interval calculated from fatigue failure of mechanical components. The proposed approach is based on the cumulative damage distribution function for evaluating mean fatigue life. Using this approach, the analytical expressions for mean and variance of the cumulative damage distribution under both stationary narrow-band and stationary wide-band random process are provided. The mean value and variance of fatigue life distribution are then evaluated to determine the optimal replacement intervals under fatigue failure. A practical example is presented to demonstrate the application of the present method.