Purpose: It is often difficult for school-based speech-language pathologists (SLPs) to prioritize implementing new practices for children with speech sound disorders (SSDs), given burgeoning caseloads and the myriad of other workload tasks. We propose that de-implementation science (e.g., Davidson et al. 2017) is equally as important as implementation science. De-implementation science is the recognition and identification of areas that are of "low-value and wasteful" (Davidson et al., 2017, p. 463). Critically, the idea of de-implementation suggests that we first remove something from a clinician's workload before requesting that they learn and implement something new. Method: Situated within the SHARE framework, we review de-implementation science and current speech-sound therapy literature to understand the mechanisms behind continuous use of practices that are no longer supported by science or legislation. We use vignettes to highlight real-life examples that clinicians may be facing in school-based settings and to provide hypothetical solutions, resources, and/ or next steps to these common challenges. Results: We identified four primary practices that can be de-implemented to make space for new evidence-based techniques and approaches: 1) over-reliance on speech sound norms for eligibility determinations; 2) the omission of phonological processing skills within evaluations; 3) homogeneity of service delivery factors; and 4) the use of only one treatment approach for all children with SSDs. Conclusions: School-based SLPs are busy, overwhelmed, and burned out (Marante & Farquharson, 2020). Although de-implementation will take work and may lead to some difficult discussions, the end result should be a reduction in SLPs' workloads and improved outcomes for children with speech sound disorders. [This paper was published in "Language, Speech, and Hearing Services in Schools" v54 n4 2023.]