The purpose of this study was to determine if there is a structure of variables that allows us to understand the level of Anxiety towards Mathematics in high school students from the municipalities of Zacatal and Jamapa, Veracruz, Mexico. This was based on the seminal works of Richardson and Suinn [1972], who developed the Mathematics Anxiety Rating Scale (MARS) instrument with 98 items. This was later modified by Alexander and Martray [1989] to develop the Revised Mathematics Anxiety Rating Scale (RMARS) with only 25 items. For this study, the test was applied to a sample of 200 high school students of first, third and fifth semesters. The reliability of the test was a: 0.934 per item and 0.693 per dimension, which suggests acceptable validity and consistency in terms of what Hair, et al. point out [1979]. For the test of the H1 and H2 hypotheses, Exploratory Factor Analysis with extraction of the Principal Components and the Chi[superscript 2][subscript c]> Chi[superscript 2][subscript t]> statistic suggest that, as a whole, the dimensions of the RMARS scale explain mathematical anxiety. In addition, they indicate that anxiety towards mathematics classes is greater than anxiety towards exams and mathematical tasks; all this accounts for 74 % of the assimilable variance. For H3, the ANOVA test is used to show if there is a difference in means. The results suggest that there are no differences by Gender, Age, Or School Grade, although the MATHTEST dimension in Gender, showed differences in variances.