Using state-level data, we report unadjusted and adjusted odds ratio of special education (SPED) trends in Tennessee from 2009-2019 for students in grades 3-8 by 3 language groups: native English speakers (NES), English-proficient bilinguals (EPB) and Current English learners (Current EL). We report trends across all SPED disability categories and across five prevalent disability categories (specific learning disability, specific language impairment, intellectual disability, other health impairments, and autism). The cross-sectional analytic sample included 812,783 students from 28 districts that met the SPED risk-ratio threshold set by the state. Results revealed that, compared to NES students, both EPB and Current EL students were generally less likely to receive SPED services, suggesting evidence of language status disparities in SPED representation. Furthermore, findings varied depending on whether adjustments were made to generate odds ratios, especially for higher-incidence disabilities (specific, learning disability, specific language impairment, and intellectual disability). Finally, the most severe evidence of underrepresentation was in lower-incidence disabilities (other health impairments and autism). Our results underscore the need for further examination into low rates of SPED identification among learners whose first language is not English (EPB and Current EL). We discuss the contextualized research, practice, and policy implications of our findings. [This paper will be published in "Journal of Learning Disabilities."]