Computer programming has been taught in secondary schools for more than two decades. However, little is known about how students learn to program. From the curriculum implementation perspectives, learning style helps address the issue of learner differences, resulting in a shift from a teacher-centred approach to a learner-focused approach. This study aims to investigate the effects of gender and learning styles on computer programming performance. The Gregorc Style Delineator (GSD) was employed to measure learning styles. A test was administered to assess students' programming performance. Two hundred and seventeen secondary school students of age from 14 to 19 participated in this study. Results indicated that no gender differences in programming performance were found after controlling for the effect of student ability. Academic ability had a differential effect on programming knowledge. Sequential learners in general performed better than random learners. These results suggest the importance of the ordering dimension of the GSD in influencing programming performance. Implications of the findings in relation to programming pedagogy are discussed in this paper.