This paper focuses on both the monitoring of the dynamic response of steel buildings with and without brace elements and the importance of model calibration on the steel buildings. The study involves the application of modal testing techniques to collect data from a three-story steel frame building model tested at the Civil Engineering Department of Karadeniz Technical University. The experimental measurements are performed under randomly generated loads. Dynamic characteristics (natural frequencies, mode shapes and modal damping ratios) obtained from bare and braced steel frame models are compared with each other. The initial analytical models of the steel model for bare and braced cases are developed and calibrated according to the experimental measurement results. The calibration process aims to minimize the differences between experimental and analytical natural frequencies. The connection rigidities of the beam-to-column are selected as a calibrating parameter in the model calibration process. It is observed that the brace elements cause an increase in the natural frequencies due to the increasing stiffness as well as the changes in the modal behavior. Static analyses of the steel frame model for bare and braced cases are carried out to emphasize the importance of the model calibration by comparing maximum lateral displacements. The calibrated analytical models produce larger lateral displacements than the initial models. The results reveal that the dynamic behavior of steel structures should be evaluated considering the calibrated models for safety of these structures. Crown Copyright (C) 2013 Published by Elsevier Ltd. All rights reserved.