Definitions of rurality in education research are inconsistent, making generalization across studies difficult at best. We review published research in rural education between 2005 and 2015 (n = 17) and characterize the way each defined rural. A common technique for classifying rural schools is the National Center for Educational Statistics (NCES) locale code. We argue that the NCES locale codes fall short of most conceptual understandings of rural. We recommend a school size filter to address the mismatch between the NCES codes and conceptual understandings of rurality. We compare non-rural districts with rural districts using NCES codes for grouping with and without the school size filter. Observed effects vary depending on whether the school size filter was used. We argue that the school size filter along with NCES codes better captures the concept of rural education and could improve research design and generalization across studies involving rural schools.