Automated Web usability analysis aims at obtaining a figure of the overall usability of a Web page or site by comparing objective page measures or indicators with statistical profiles or other forms of pre-existing knowledge. Existing automated analysis tools based on empirical data make use of correlations between indicators and expert judgments, but do not address the modeling of each individual concept describing characteristics relevant to usability, as those that commonly are used to formulate usability guidelines. In this paper, we describe how such concepts can be modeled through experimental design, resulting in a novel form of usability analysis tools combining empirical data and design guidelines. As vagueness is an inherent characteristic of perceptual concepts related to usability, fuzzy set elicitation techniques are used to model the concepts, and fuzzy rules are used to combine them in flexible decision procedures. KEYWORDS Web usability analysis, usability guidelines, ...