As the Internet has made virtually unlimited resources available to people to learn on their own, more and more people are engaging in self-directed learning to enhance their educations. Researchers have investigated self-directed learning for years, focusing on variables such as design of content and learner attributes such as motivation and self-efficacy as predictors of learning success. The present paper looks at student aptitude as a predictor of student performance in a self-directed learning context. 33 middle and high school students, who were or had been in either gifted and talented or mainstream educational programs in school were taught how to program websites. They were either taught by a teacher or by watching videos (the self-directed learning condition). At the end of the instructional period, all students were asked to create a website about themselves and the quality of their work was graded by experienced computer programmers. Results showed that gifted and talented students performed equally well with or without a teacher while non-gifted and talented students performed at a level equal to the gifted and talented students when they had a teacher but lower when they did not. Results suggest that self-directed learning may be equally effective as teacher-led education for gifted and talented students, but non-gifted and talented students may benefit more from traditional education than self-directed learning.