Research has shown that views are an important factor in the overall satisfaction of building occupants, but, because they are qualitative and subjective, views are difficult to assess. This paper presents a new approach to analyzing views using a fuzzy logic methodology, with the aim of assessing the amenity of views. It illustrates the process of constructing a system from data collection and fuzzy rules design, to inference and defuzzification procedures. With the verification of the prediction outcomes, it is shown that the evaluation model can successfully mimic a human's subjective preferences for scenic views. The methodologies and the underlying principles of the evaluation model are explained.