Two studies were conducted to compare subjects' performance reading texts displayed on a computer terminal screen and on paper. In the first study, 10 graduate students read a 1,000-word article on knee injuries from "Science 83" magazine and were tested for recall of information on eight items. While subjects in the control condition (reading hard copy) were more accurate in spatial recall of text sequence, vertical spacing, and horizontal spacing, the results were statistically significant only in the category of vertical location. In the second study, 15 graduate students were timed for retrieval of information. Subjects read texts in three conditions: (1) hard copy (paper), (2) a standard CRT screen with text presented on 12 screens with a baud rate of 4800, and (3) a 19-inch diagonal screen with text presented in 5.5 screens. Results showed that subjects read significantly faster under conditions one (paper copy) and three (large screen) than under condition two (normal CRT.) (Author/HOD)