A innovative macroseismic approach allowing the vulnerability analysis of building typologies from poor data, is presented in this paper. The building typologies defined within the European Macroseismic scale (EMS-98) have been considered. The Damage Probability Matrixes have been derived interpreting the correlation to the vulnerability classes suggested by EMS-98 in terms of relative frequencies associated with the vulnerability classes. Bayes theorem allows the upgrade of the frequencies when further data about the built-environment or specific properties of the buildings are available, allowing the identification of a different behaviours on respect to the one generally considered for the typology. Fuzzy measures of any damage function can be derived, using parametric or non-parametric damage probability matrices implicitly contained within EMS-98 scale. For every result of the seismic analysis, the procedure allows supplying to the user the final uncertainty connected with the aforementioned fuzzy relation between the probability of the damage grade, the macroseimic intensity and the vulnerability class Numerical implementation based on ISTAT 1991 data are presented for a set of 19 Municipalities that can be regarded as representative for a wider area in the Veneto and Friuli Regions. The measures are related to the expected percentage of inhabitable and collapsed building for a fixed and uniform value of the Macroseismic Intensity in the area.