Research directed toward a better understanding of the computer user/computer machine relationship has increased in recent years. To identify what factors may predict success in computer programming, 286 college students from three computer classes (160 from introductory programming; 60 from Cobol programming; and 66 from senior level programming) completed a 30-minute question answering and problem solving test. Cognitive factors (past academic performance and problem solving ability) were assessed through self-reported grades and seven math and logic problems. Personality factors (computer anxiety and locus of control) were assessed by items from the Fennema and Sherman Mathematics Attitude Scale and a modified internal/external (I-E) scale of personal locus of control. An analysis of the results showed that past academic performance in mathematics and English courses, amount of previous computer experience, expected grade in the course, and performance on selected logic and algebraic-word problems were significantly correlated to course performance. Neither of the personality factors predicted performance, since few students showed any computer anxiety and most showed an internal locus of control. (BL)