AbstractIn the present work, dynamic analyses are performed on a footbridge considering deterministic and probabilistic models with two distinct formulations of a human walking loading simulation, a force model (FM) and a biodynamic model (BM). The probabilistic analyses are based on the random walk (RW) method in order to realistically represent the pedestrian movement with interperson and intraperson variability approaches. For the methodology’s development, a theoretical-experimental correlation is performed based on tests of one and six people crossing a steel footbridge. A performance assessment is carried out following current guidelines along with a statistical analysis using RW. The study aims to provide a better understanding of the structure in serviceability scenarios using the two probabilistic RW loading models with FM and BM (RWFM and RWBM). The results show that most of the maximum acceleration amplitudes reaches the unacceptable comfort limit using the RWFM, and the comfort level is improved in the simulations using the RWBM. The results obtained in the proposed simulations point out that the use of RW successfully simulates the slight variations present in the walking cycle.