English Language Proficiency Assessments: The Complete Guide

Diverse English learner classroom, English language proficiency assessments guide

Estimated reading time: 8 minutes

An English language proficiency assessment is the annual, standards-based test a state uses to identify English Learners and measure their progress toward full proficiency. If you have searched for an English language proficiency test and landed on a different acronym in every state, that is because there is no single national exam: instead there are several, each used by a group of states. This guide maps the major systems, WIDA ACCESS, ELPA21, ELPAC, TELPAS, NYSESLAT, and AZELLA, explains what they share, and links to a detailed guide for each.

Table of contents

Executive Summary

Every state is federally required to identify English Learners and measure their English language proficiency each year, but states do not use the same test. Most use one of two large consortium assessments, WIDA ACCESS or ELPA21, while several large states run their own: California uses ELPAC, Texas uses TELPAS, New York uses NYSESLAT, and Arizona uses AZELLA. Despite different names and scales, all of them assess the same four language domains, use a screener for identification and a summative test for annual progress, and tie exit from services to reaching proficiency. Understanding the shared structure makes any state’s system easier to read.

Key Takeaways

What Is an English Language Proficiency Test?

An English language proficiency test is a standards-based assessment that measures how well a multilingual student understands and uses academic English across four domains: listening, reading, writing, and speaking. Federal law requires states to identify English Learners and assess their proficiency annually, which is why every public district administers one of these tests. Each system has two parts: a screener given at enrollment to identify students who need services, and a summative test given each year to measure progress and signal when a student is ready to exit.

The Major English Language Proficiency Assessments

The table below maps the main systems used in U.S. K-12 schools. For depth on any one, follow its linked guide.

AssessmentWhere usedIdentification screenerProficiency scale
WIDA ACCESSRoughly 40 states and territoriesWIDA Screener6 levels (Entering to Reaching)
ELPA21Nine statesDynamic Screener5 levels per domain; overall Emerging, Progressing, Proficient
ELPACCaliforniaInitial ELPAC4 performance levels
TELPASTexasTELPAS identification4 levels (Beginning, Intermediate, Advanced, Advanced High)
NYSESLATNew YorkNYSITELL5 levels (Entering to Commanding)
AZELLAArizonaAZELLA placementProficiency levels (Pre-Emergent to Proficient)

Detailed guides: WIDA, ELPA21, ELPAC (California), TELPAS (Texas), NYSESLAT (New York).

What They All Share

The acronyms differ, but the architecture is remarkably consistent. Every system in the table above does four things the same way.

Two Consortia and the State-Specific Tests

Most states belong to one of two consortia. WIDA, based at the University of Wisconsin-Madison, is the larger, and its ACCESS for ELLs test is the most widely used English language proficiency assessment in the country. ELPA21, housed at UCLA’s CRESST, serves nine states with a comparable structure. The remaining large states built their own systems to align tightly with their state standards: California’s ELPAC, Texas’s TELPAS, New York’s NYSESLAT, and Arizona’s AZELLA. The result is a patchwork, but a legible one once you know the shared parts. Our detailed comparison of ELPA21 vs WIDA shows how two of these systems line up.

Why Scores Do Not Transfer Between States

Because each system uses its own scale and each state sets its own reclassification rule, a proficiency result does not carry across state lines. A student who is one level short of exiting in a WIDA state is generally re-assessed under the new state’s system after a move, not simply converted. For district leaders and mobile families alike, the practical rule is to read proficiency in the terms of the state you are in, and to expect a fresh evaluation when a student crosses a state boundary.

District Benchmark

Whatever the test, the arithmetic of an EL program looks similar. A 5,000-student district with 12 percent English Learners serves about 600 students who take a proficiency test every spring, plus a stream of new arrivals who must be screened at enrollment. The system’s name changes the paperwork and the score labels, but not the core task: moving each of those students toward proficiency in all four domains, and identifying which domain is holding each one back. Across every system in this guide, that domain is most often speaking.

The Shared Speaking Bottleneck

The most useful thing to know about all of these assessments is what their results have in common. In WIDA states and ELPA21 states, in California and Texas and New York, students reach proficiency in listening and reading before writing and speaking, and speaking is the domain that lags the longest. The reason is the same everywhere: speaking proficiency grows through many short, responsive conversations, and one teacher cannot hold enough of them with a full class to give every student that volume of practice. The test a state chooses is a policy detail; the speaking bottleneck it reveals is universal. Our guide to EL speaking practice covers how districts build more of the practice every one of these systems rewards.

Funding Options

Assessment preparation, EL services, and supporting technology are allowable under the same federal and state funding across every state, regardless of which test it uses.

Funding SourceEligible UsesHow to Access
Title IIISupplemental EL instruction, technology, professional developmentFormula grant through your state DOE
Title IAcademic support in qualifying schoolsFormula grant through your state DOE
State EL fundingVaries by state weighted formulaCheck your state DOE website
TIPS CooperativePre-approved vendor purchasingNo separate RFP required via mytelo.ai

See our complete guide to Title III funding.

Frequently Asked Questions

What is an English language proficiency test?

It is the annual, standards-based assessment a state uses to measure how well an English Learner understands and uses academic English across listening, reading, writing, and speaking, and to decide when the student can exit language services.

What are the main English language proficiency assessments?

The two large consortium tests are WIDA ACCESS and ELPA21. Several states use their own: ELPAC in California, TELPAS in Texas, NYSESLAT in New York, and AZELLA in Arizona.

Which English language proficiency test does my state use?

Most states use WIDA ACCESS; nine use ELPA21; California, Texas, New York, and Arizona use their own. Confirm with your state department of education, since membership can change.

Do all the tests measure the same skills?

Yes. Every major system assesses the same four domains, listening, reading, writing, and speaking, and pairs a screener at enrollment with an annual summative test.

Do proficiency scores transfer between states?

Not directly. Scales and reclassification thresholds differ by state, so a student who changes states is generally re-evaluated under the new state’s system.

Conclusion

The map of English language proficiency assessments looks fragmented, but the systems rhyme: same four domains, same screener-and-summative structure, same proficiency-gated exit, same speaking bottleneck at the end. A district leader who understands that shared architecture can read any state’s test, and can focus energy where it actually pays off, on the practice that moves students toward proficiency in every system that measures it.

Sources: WIDA at the University of Wisconsin-Madison; CRESST, ELPA21 at a Glance; U.S. Department of Education, NCELA.

The One Thing Every Scale Has in Common

Line up all six systems and the surface differences are obvious: six levels here, four there, a composite score in one state and a three-part determination in another. Look at what the results actually show, though, and the differences fall away. In every system, the same domain arrives last, and for the same reason.

Speaking is the skill a classroom can least manufacture. It grows through individual conversation, and a single teacher divided across a full class can give each student only minutes of it a day, far short of what any of these scales rewards at the top. That is why speaking decides reclassification timelines from California to West Virginia. Districts that move students fastest, whatever their state’s test, tend to solve for that one constraint directly, giving each learner individualized speaking practice alongside instruction. Telo AI is one example, and you can see how the practice works and whether it fits your program. The assessment is set by your state; the practice that carries students to the top of it is set by you.