Abstract : In this work fragility analysis of reinforced concrete and steel structures withinfill walls is performed.For this purpose a fuzzy-based fragility assessment framework for evaluating 3D framed structures is proposed taking into account various sources of uncertainty. In particular, randomness on the material properties and on the seismic demand is consid-ered. The proposed framework requires the development of a fuzzy nonlinear static analysis model in order to define the limit states. The fragility curves are expressed in the form of a two-parameter lognormal distribution. Keywords: Fuzzy response, fragility analysis, fibre modelling, nonlinear static analysis, steel and reinforced concrete struc-tures. 1. INTRODUCTION Seismic fragility analysis provides a measure of the safety margin for the structural system, sincefragilities repre-sent the probabilities of exceedance of limit-states as func-tions of earthquake ground motion intensity. Therefore, fra-gility analysisis considered as the main ingredient of the risk assessment procedure.In the past a number of studies on fra-gility analysis of structural systems have been published[1-5]. Real-world structures are characterized by imperfections while the material properties and the loading conditions are uncertain, which induce deviations from the nominal state assumed by the design codes. A deterministic representation of a design that ignores scatter of any kind of the parameters affecting its response is never materialized in an absolute way, due to unavoidable scattering of the values of its parameters. So far a number of researchers studied the effect of uncertainties in the context of fragility analysis; mainly in steel and RC structures [6-10]. Modelling uncertainty in engineeringproblems as random variables or random processes becomes problematic when uncertain data have additional uncertainty besides the prop-erty of randomness. Fuzzy randomness is a generalized un-certainty model to describe samples with uncertainty of the single sample element. The basic terms and definitions re-lated to fuzzy randomness can be found in [11-13]. Recently, non-probabilistic approaches for numerical engineering problems with uncertain variables have been proposed [14,15], such as convex models, fuzzy sets, random sets, possibility the oryand others. The main motivation for adopt-ing nonprobabilistic approaches is the high sensitivity offail-ure reliability to the tails of probability distributionsof the random variables involved in the analysis [16]. Publications implementing fuzzy numbers in fragility analysis are limited max