This paper presents modal parameter estimation work on a steel railway bridge at three different temperature conditions using ambient vibration test data. The bridge was built at the end of the 19th century using the available technology of its time. It is composed of six spans, each 30 m long, with a total length of 180 m. It is slightly curved in the horizontal plane with a radius of 300 m, and has a vertical grade of 2.5%. Modal parameters of the bridge were estimated using two different output-only system identification methods. The identified results obtained under different temperature conditions were compared in assessing the effects of temperature variation in the identification results. A comparative study in assessing method-to-method and test-to-test variability was also conducted. A three-dimensional finite-element model of the bridge was developed. In order to match the experimentally obtained modal parameters with the numerical ones, a trial-and-error–based model updating study was conducted. This way, a benchmark model of the 199+325 steel railway bridge was obtained for future capacity assessment, prediction, and sensitivity-based model updating work.