Several alternative methods of presenting written material within a graphic were evaluated to provide empirical data on how technical writers and artists might improve the comprehensibility of graphics. The number, location, and sequence of "callouts"--the pictorial part references using arrows--were systematically varied in a series of drawings and tested on samples of Navy trainees. The major finding indicated that arranging callouts sequentially provides for good comprehension, even when the number of callouts is large. It is recommended that clarification of the graphic comprehension issue be pursued through empirical studies of users' information-search behavior and the stimulus variations that influence its effectiveness. (Author/CMV)