Bridge Health Monitoring (BHM) under realistic traffic and noise conditions remains a critical challenge due to the inherent variability in vehicle parameters and measurement uncertainties. This paper assesses damage detection reliability through Probability of Detection (PoD) curves, a statistical tool widely used in Non-Destructive Testing (NDT) but rarely applied to traffic-induced bridge responses. The application of the PoD curve concept to bridge responses from a test vehicle with known properties was first introduced by Li et al (2025) and is extended here to a traffic fleet of 2-axle trucks with random properties. A 2-axle truck is simulated using a half-car model, and the bridge is represented as a 15 m simply supported finite element beam model. The stiffness of the beam is assumed constant throughout the span and decreases uniformly as a result of global damage. The traffic fleet is modelled by adopting statistical distributions for gross vehicle weight, axle weights and spacing, speed, lateral position, and suspension properties of the half-car, based on literature and real data gathered from a Weigh-In-Motion site in France. Monte Carlo simulations are employed to generate the response of the bridge to multiple crossings of 2-axle trucks randomly sampled from the assumed distributions. These simulations incorporate additive Gaussian noise and an ISO class ‘A’ road profile to replicate field conditions. The bridge responses investigated in this paper are midspan displacement, midspan strain, and rotation at the support, all of which could potentially be recorded by a BHM system. For each truck crossing, the difference between the bridge’s response at original stiffness and reduced stiffness is computed, and when the variation exceeds a pre-established threshold, the change in stiffness is deemed detectable. This threshold is directly related to the false alarm rate, with the proportion of responses exceeding it determining the PoD for a given stiffness loss. By integrating BHM data, stochastic fleet dynamics, and PoD analysis, this paper provides actionable detection criteria for engineers as a function of measured response and noise level, enhancing decision-making in bridge management. References.- Li, C., González, A. and Casero, M. (2025). Probabilistic assessment of the sensitivity of direct and drive-by measurements to a global stiffness loss in a bridge traversed by a vehicle. Available at SSRN: http://dx.doi.org/10.2139/ssrn.5251785