Prior research has shown that students from low socio-economic status (SES) families are at greater risk for mathematics education and achievement, and these factors in turn, may impact their long-term well-being. This paper investigates differences in mathematics achievement by SES, using clinical interview (CI). Students in Kindergarten to Grade 3 were studied using cross-sectional and longitudinal data to examine differences in mathematics accuracy, strategy use, and profiles. Results indicate minor differences in overall accuracy and in overall profile scores. At the subtest level, differences were found for accuracy for all grade levels. There were significant differences in mathematics performance by SES. In particular, larger differences by SES were found in mathematics accuracy than in sensible strategy use. For profile scores, significant differences were found by SES using cross sectional data; moreover, students' profile scores differed by subtests and sections of the CI. Longitudinal data analysis indicated significant changes in students' profile scores; however, there were no differences in profile scores by SES over time. Curricular and assessment implications are discussed, with directions for future research.