This paper examines the dynamic characteristics of a steel arch footbridge and uses Bayesian model updating to identify uncertain input parameters. The modal parameters of the bridge (frequencies and mode shapes) were determined through operational modal analysis (OMA) and a finite element model was created using ANSYS. The experimental findings and the numerical results were compared using the modal assurance criterion to identify the pairs of aligned modes. A one-at-a-time sensitivity analysis identified the most influential input parameters on the response of the FE model. Seven input parameters were selected for Bayesian model updating, modeled as beta random variables. The likelihood function and posterior distribution were evaluated using the Markov chain Monte Carlo (MCMC) method with a surrogate model based on general polynomial chaos expansion (gPCE).