Structural reliability techniques can be employed to evaluate the fatigue performance of fracture-critical members in steel bridges. In this dissertation, two fatigue reliability formulations that can be applied for most details in steel bridges are developed. For details classified according to AASHTO fatigue categories, a limit state function related to the number of stress cycles leading to failure based on Miner’s rule is used; for details not classified according to AASHTO fatigue categories, a limit state function based on linear elastic fracture mechanics and expressed in terms of crack size and growth rate is employed. With the application of fatigue reliability analysis, a procedure for inspection scheduling of steel bridges is developed to yield the optimal (most economical) inspection strategy that meets an acceptable safety level through the planned service life. This inspection scheduling problem is modeled as an optimization problem with an objective function that includes the total expected cost of inspection, repair, and failure formulated using an event tree approach, with appropriate constraints on the interval between inspections, and a specified minimum acceptable (target) safety level. With the help of several illustrations, it is shown that an optimal inspection scheduling plan can thus be developed for any specified fatigue details or fracture-critical sections in steel bridges. A second optimal inspection scheduling procedure is formulated that takes into consideration crack detectability (or quality) of alternative nondestructive inspection techniques. This procedure based on Monte Carlo simulation of crack growth curves yields an optimal inspection technique and associated schedule for a given fracture-critical member in a steel bridge for minimum cost and a target safety level while also taking into account probability of detection (POD) data for candidate nondestructive inspection techniques. Comparisons between the reliability-based procedure and the POD-based procedure for optimal inspection scheduling are discussed. Both scheduling strategies, when contrasted with ad hoc periodic inspection programs for steel bridges, are recommended because they are rational approaches that consider the actual fatigue reliability of the bridge member and account for economy as well as safety. / text