This study used structural equation model to determine the relationship of personal background, school environment, and family background to public and private secondary school graduates’ performance in high stakes test. Results revealed that for public school graduates, study hours in school, library use, and sib size are all significantly and negatively related to their high stakes test results, indicating that those who got higher scores in high stakes test spent lesser study time in school, and had lesser usage of library. Additionally, those who came from smaller family also performed better. For private school graduates, reading activity and extra-curricular activity are significantly and positively related to their high stakes test results while the relationship of library use is negative. Graduates from urban places performed better and those who belong to bigger class size still obtained higher scores. Moreover, those who belong to lower economic status and with lower parental involvement got higher scores. Based from the results, this study revealed that higher library usage cannot explain students’ performance. This is because, information can be easily accessed using the latest educational technology anytime and anywhere. However, schools still need to take necessary actions to improve the library facilities and make it more attractive and useful to the students.