This paper presents an Artificial Intelligence approach towards classification of persons based on a verbal description of their facial features. Face features are extracted by use of an existing detection techniques, such as measurements of horizontal and vertical size of the facial elements like nose, eyes, etc. This approach allows to create fuzzy sets representing selected facial features. What is important, linguistic variables corresponding to the fuzzy sets conform the terminology applied by law enforcement to create an eyewitness verbal description. Then, fuzzy IF-THEN rules are employed for classification both the facial composites (sketches) and usual images from face databases. With regard to this concept, satisfactory results have been obtained and presented.