The U.S. Bilingual Education Act (BEA) of 1968 was born of grassroots efforts by linguistically diverse communities and their allies. Advocates' goal was to achieve self-determination in the education system and beyond. However, the BEA was implemented as a title under the Elementary and Secondary Education Act, which mandated annual program evaluations. Stipulations for these evaluations were strict, and they evolved over time to require the use of standardized assessments so that program outcomes could be compared. Many local evaluators and the communities they served objected to these requirements on the grounds that existing standardized assessments were inappropriate for assessing the bilingual program; however, most evaluators did comply with the mandate. The goal of comparability remained elusive, so the U.S. Office of Education contracted for a national impact study. This study's results were controversial and became a critical juncture in federal policy for linguistically diverse students. Using a Foucauldian frame, I draw on evaluative reports from the field, federal statutes, congressional debates, and federal program guidelines to investigate the roles of evaluation policy and practices during the first decade of the BEA.