The background of 197 university students from four introductory computer science classes at the State University of New York at Buffalo was investigated to determine factors contributing to academic success or failure in computer science. A background questionnaire, a test of computer science concepts (reading a simple program, symbol manipulation, language constructs, and algorithmic processing), and a learning style inventory were administered. It was hypothesized that the study would find: (1) significant differences in achievement for those who have started the study of computing prior to high school, in high school, and at the college level (rejected); (2) a significant positive correlation between mathematics scores and self-perceived prior mathematics achievement and current achievement (accepted); (3) positive correlations between achievement and both the number of semesters of computer studies (rejected) and the length of programs written (accepted); (4) significant differences in achievement between students with different initial programming language experience (rejected); (5) significant differences in learning style preference and achievement (rejected); and (6) significant differences in learning style preference for students selecting computer science as a major (accepted). Error analysis was performed on responses, and test scores were compared with sources of learning. It was concluded that the greatest positive impact on achievement derives from early exposure to computers and length of programs written. (98 references). (MES)