Tests, assessments, and evaluation instruments have long been ill-equipped to account practices of translanguaging inherent within multilingual communities (Author, 2019; Duran, 1989; Sánchez, 1932). Within the field of testing and assessment, more emphasis has been placed on ensuring design principles of the assessment. In making adaptations to existing monolingual content tests, for example, test developers have focused on reducing construct-irrelevant variance of test performance due to bi-/multilingualism or mitigating potential “differential boost” that where a bi-/multilingual person may perform better on an assessment than a monolingual person. Though presented as connected with fairness and equity, in practice these approaches further add to the long-standing lineage of assessment work that has sustained deficit perspectives of bi-/multilingual peoples (Author, 2020). Countering over 100 years of deficit positioning of bi-/multilingualism in testing recently has emerged in work that values the knowledge, beliefs, and experiences of teachers (Ascenzi-Moreno, 2018). This participatory action research project examines how four Spanish/English bilingual teachers working in elementary schools in the United States integrate knowledge and expertise to critique existing assessment design principles and practices during a 5-week online course on classroom assessment. Transcripts from online meetings, individual work samples re-envisioning assessment design, and data from a peer and individual meta-test critique serve as our key data sources. All authors on the paper were also participants in the course and have provided key insights throughout the research project from data analysis to dissemination. Although our findings report an overall position that supports the use of translanguaging in assessments, we also delve into the structural constraints, in particular, the move for most schools to have computer-generated—rather than teacher-created—assessments.