Assuring that English language learners (ELLs) receive the services to which they have a right requires accurately identifying those students. Virtually all states identify ELLs in a two-step process. First, parents fill out a home language survey. Second, students in whose homes a language other than English is spoken and who therefore might be less than fully proficient in English, are tested for English language proficiency. The home language survey thus plays a gatekeeping role. If it fails to identify potential ELLs, there is a greatly reduced chance these students will be identified and receive services to which they are entitled. The two studies reported in this paper are specifically about the process whereby potential ELLs are identified so that they might be tested, then receive services if they qualify. More specifically, it addresses the question of whether Arizona's sharp reduction in the home language survey questions can lead to failure to identify students who, by the state's own criterion (i.e., performance on the AZELLA), are entitled to those services. Analyses of data from two Arizona school districts clearly show that use of a single home language survey question will under-identify students. Based on data from these two districts, as many as 11 to 18% of students who are eligible for ELL designation could be denied services to which they are entitled if a single home language survey question is used to identify potential ELLs. Further, it is highly unlikely that a fail-safe mechanism established by the state, whereby teachers can nominate potential ELLs for language testing, will in fact successfully identify most students the new procedure fails to identify. (Contains 4 tables and 4 footnotes.) [Abstract was slightly altered to meet ERIC standards.]