School-based learning experiences are often designed with the "typical" student in mind. However, this may not be an optimal approach, given the variability of prior learning that exists in most classrooms. We investigated the variance in achievement within U.S. fourth- and eighth-grade mathematics classrooms using Trends in International Mathematics and Science Study (TIMSS), an international assessment that uses representative samples within participating countries and has pre-established benchmarks. Approximately 23% of students in a typical Grade 4 classroom were found to score at or below the low benchmark, whereas 14% met or exceeded the advanced benchmark; these numbers are 35% and 14% for Grade 8 classrooms. That is, nearly 38% and 49% of students in the typical Grade 4 and 8 classroom, respectively, may require instruction substantially different from typical "grade-level" instruction to allow them to work within their zone of proximal development for mathematics. Furthermore, approximately 69% of Grade 4 and 35% of Grade 8 classrooms included students across all four benchmarks. Finally, about 68% and 37% of the variance lay within classrooms at each grade level, respectively. These results provide evidence that one-size-fits-all age-based approaches to instruction are unlikely to provide effectively targeted learning experiences for all students. Instead, a variety of services, such as flexible ability grouping or acceleration, may provide more relevant instruction.