Reasons for students' self-perceptions of mathematical talent form the basis for this predominantly qualitative study. Participants (N=60) were selected from a larger cohort of 459 Year 9 students from three coeducational government schools in an upper-middle class area of metropolitan Sydney, of comparable socioeconomic status. Interviewees were selected from the large-scale extensive study according to their perceived mathematical talent self-ratings, and their measured mathematical performance on standardised tests. Interview questions focused on the self, significant others, and wider sociocultural influences. Focal groups were girls of high mathematical performance with high self-perceptions of talent, boys of low performance and high talent perceptions, and girls of high performance and low talent perceptions. These groups were of interest in order to explore factors facilitative of high-achieving girls' corresponding high talent perceptions, factors contributing to boys' inflated talent perceptions, and most importantly, factors that act as deterrents to high-achieving girls having high talent perceptions. Cross-sex parallel comparison groups were included in each cell in order to enable identification of gender-specific processes, resulting in a total of six groups, each comprising ten interviewees. Content analysis identified emergent themes for each gender, level of talent perceptions, level of performance, and interactions among these.