This descriptive research is grounded on the objectivist epistemology and informed by positivism. It used Path Analysis to examine the predictive and mediational role that creativity, self-efficacy, and anxiety play in the mathematical problem solving performance of potential mathematically gifted Grade Six pupils from selected elementary schools in Iloilo. The eighty-three participants were given the Kuhlmann-Anderson Test to determine their cognitive ability. The results of the test were also used as the basis for their classification into High Potential Mathematically Gifted (HPMG) and Low Potential Mathematically Gifted (LPMG). Based on the results, 40 pupils were classified as HPMG while 43 as LPMG. The instruments used to gather data were the Mathematics Anxiety Rating Scale (MARS), Mathematical Creativity Test (MCT), Mathematics Self-Efficacy Rating Scale (MSRS), Mathematical Problem-Solving Test (MPS), and Parental Support Rating Scale (PSRS). Means, standard deviations, and percentages were used for descriptive data analyses and the Pearson’s Product-Moment Correlation, Multivariate Analysis of Variance (MANOVA), Multiple Linear Regression, and Path Analysis (PA) for inferential data analyses, all set at .05 level of significance. Results showed that as a whole group, the participants reported a moderately high cognitive ability and self-efficacy, low anxiety; and average creativity and problem-solving performance. MANOVA revealed a statistically significant multivariate effect between the high and low potential mathematically gifted groups on the combined dependent variables. In the final path model, cognitive ability significantly influenced all endogenous variables; parental support predicted self-efficacy and anxiety; self-efficacy predicted anxiety and creativity; and creativity predicted problem-solving performance. The total effect of cognitive ability, sex, and parental support on problem-solving performance suggests that part of their influence was mediated by pupils’ self-efficacy perceptions, anxiety, and creativity. Self-efficacy, anxiety, and creativity, on the other hand, mediated the effect of cognitive ability and parental support as their respective total effects on problem-solving performance were much stronger than their direct effects. It is inferred, then, that a potential mathematically gifted pupils’ innate ability, if coupled with high sense of mathematics efficacy, and ability to produce many ideas, to generate varied approaches observed in a solution, and to come up with novel and unique ideas will make them a successful problem solver.This descriptive research is grounded on the objectivist epistemology and informed by positivism. It used Path Analysis to examine the predictive and mediational role that creativity, self-efficacy, and anxiety play in the mathematical problem solving performance of potential mathematically gifted Grade Six pupils from selected elementary schools in Iloilo. The eighty-three participants were given the Kuhlmann-Anderson Test to determine their cognitive ability. The results of the test were also used as the basis for their classification into High Potential Mathematically Gifted (HPMG) and Low Potential Mathematically Gifted (LPMG). Based on the results, 40 pupils were classified as HPMG while 43 as LPMG. The instruments used to gather data were the Mathematics Anxiety Rating Scale (MARS), Mathematical Creativity Test (MCT), Mathematics Self-Efficacy Rating Scale (MSRS), Mathematical Problem-Solving Test (MPS), and Parental Support Rating Scale (PSRS). Means, standard deviations, and percentages were used for descriptive data analyses and the Pearson’s Product-Moment Correlation, Multivariate Analysis of Variance (MANOVA), Multiple Linear Regression, and Path Analysis (PA) for inferential data analyses, all set at .05 level of significance. Results showed that as a whole group, the participants reported a moderately high cognitive ability and self-efficacy, low anxiety; and average creativity and problem-solving performance. MANOVA revealed a statistically significant multivariate effect between the high and low potential mathematically gifted groups on the combined dependent variables. In the final path model, cognitive ability significantly influenced all endogenous variables; parental support predicted self-efficacy and anxiety; self-efficacy predicted anxiety and creativity; and creativity predicted problem-solving performance. The total effect of cognitive ability, sex, and parental support on problem-solving performance suggests that part of their influence was mediated by pupils’ self-efficacy perceptions, anxiety, and creativity. Self-efficacy, anxiety, and creativity, on the other hand, mediated the effect of cognitive ability and parental support as their respective total effects on problem-solving performance were much stronger than their direct effects. It is inferred, then, that a potential mathematically gifted pupils’ innate ability, if coupled with high sense of mathematics efficacy, and ability to produce many ideas, to generate varied approaches observed in a solution, and to come up with novel and unique ideas will make them a successful problem solver.