A study on visual literacy in children explored such questions as: (1) how a visual literacy curriculum might be structured; and (2) whether, when students participate in a visual literacy program, they consequently say or write more, and whether what they have to say is more insightful. During the experiences, children talked and wrote about pictures in books after being introduced to such visual elements as line, shape, color, texture, form, and space. Results reported are based on data from 15 fifth grade children from each of two classrooms, one suburban and one urban. Subjects were five children each in high, medium and low language groups. Students were asked to apply some of what they were learning about visual elements to a very concrete task: choosing a Caldecott winner and giving reasons for their choices. To analyze the data reported, three different language measures were used, two quantity measures and one quality measure. Results indicated the number of words generated in both lessons was about the same. In each lesson, the urban children generated more words than did the suburban children. When the categories of comments are further grouped into those relating specifically to pictures, and those related to factors other than the pictures, it was clear that children in these two groups had become able to talk about visual reasons for preferring one book over another. (Eight tables of data are included.) (MG)