Predictors of Mathematics Performance: An Inductive Approach Towards Theory Generation. Objectives: This study aimed to examine the predictive factors that affect the mathematics performance of the students towards developing a theory. Methods: Quantitative non-experimental method employing correlational approach was used in the study and inductive approach as basis for the theory development. Multiple regression was the main analysis in this study. Random sample students from the English and Communication Department participated in the study. The researcher used three standardized questionnaires, Mathematics and Technology Attitudes Scale (MTAS) and Revised Mathematics Anxiety Rating Scale (RMARS). Findings: The findings exposed that of the 15 domains, effective communication has the highest mean while self-efficacy has the lowest. This study also discovered that there was a significant difference between the sex and students’ level of anxiety in terms of numerical test anxiety and mathematics course anxiety. The multiple regression also further revealed that of the 15 domains, only instructional planning strategies and effective communication were the best predictors of mathematics performance with regression equations of Y= 1.563 + 0.358X and Y= 1.563 – 0.556X, respectively. Conclusion: This study concluded The level of mathematics performance of the students is multi-factorial in nature which depends on teachers’ delivery, pedagogies, and coping with assessment anxiety. Keywords: mathematics performance, predictive factors, students’ attitude, teachers’ effectiveness, theory generationDOI: http://dx.doi.org/10.23960/jpp.v13.i1.202303