A statistical model is developed to estimate the seismic vulnerability of low- to mid-rise reinforced concrete buildings. The model is based on a novel utilization of the discriminant analysis technique of multivariate statistics. Number of stories above the ground level (N), soft story index (SSI), overhang ratio (OHR), minimum normalized lateral stiffness index (MNLSTFI), minimum normalized lateral strength index (MNLSI) and normalized redundancy score (NRS) are selected as the basic estimation variables. The earthquake damage data compiled for the 12 November 1999 Duzce earthquake are used to develop a discriminant function in terms of these estimation variables. The discriminant score obtained from the resulting discriminant function is then used to estimate the damage state of buildings ranging from no damage to collapse, with intermediate damage states of light, moderate and severe. Correct classification rates ranging between 62% and 95% obtained for the seismic damage data associated with the recent earthquakes that occurred in Turkey support the predictive ability of the proposed model. (C) 2003 Elsevier Ltd. All rights reserved.