English proficiency in Latin America by country (2026): EF EPI scores and how to test candidates

Argentina, by a clear margin: 575 points on the EF English Proficiency Index 2025 and the only Latin American country in the top 30. Honduras (553, 32nd) is the only other one in EF’s high band. Uruguay follows at 542, and twelve countries sit in the moderate band between 500 and 549. At the bottom of the region are Brazil (482), Colombia (480), Ecuador (466), Haiti (444) and Mexico (440).[5]

The band column applies EF’s own score ranges to each country’s score.[6]

EF groups scores into five bands and maps them to levels of the Common European Framework of Reference (CEFR), with examples of what a person at that level can do.[6] For hiring, the useful line is between moderate and high: moderate means someone can take part in meetings in their area of expertise and write professional emails on familiar subjects, while high adds making a presentation at work.

For a role that talks to US customers all day, such as sales, account management or support, aim for candidates who perform at the high or very high level in your own interview. For a role that works mostly in writing with an internal team, the moderate level can be enough if the written test is strong.

Useful for comparing countries, weak for predicting a candidate. The 2025 index is built from more than 2.2 million people who took EF’s online, adaptive test (the EF SET) in 2024, and EF states that this test-taking population is self-selected and not guaranteed to be representative.[6] Because the test is online, people without internet access are left out, and national scores are three-year rolling averages.[6]

Internet use also varies: among the four largest hiring markets it ranged from 89.7% of the population in Argentina to 79.3% in Colombia in 2024,[23] so the people behind each score are not the same slice of society everywhere. Read the index as a description of adults who chose to take an English test, not of the people who apply to your job.

Brazil gained the most in the 2025 edition (+16 points), followed by Argentina (+13), Haiti (+12), El Salvador and Venezuela (+10 each). The largest falls were Mexico (−19) and Costa Rica (−18), with Chile (−8) further behind.[5]

Because each national score is a three-year rolling average of a self-selected sample,[6] read a single year’s change as a signal to watch rather than a trend.

No. A country average blends millions of people, and the professionals who apply for a US remote role are a small, self-selected group within it. Mexico ranks lowest in the region, yet its labor force is 61.7 million people, and Brazil’s is 108.3 million;[24] even a small fluent minority is a large pool in absolute terms. Use the country score to set expectations for how long a search may take, not to rule out a country.

Test the English the job uses, in the format the job uses it. A short protocol that works for most roles:

1. Hold the first interview in English, on video, and include at least one question that requires an explanation rather than a rehearsed answer.
2. Add a written task in the language of the role: a customer email, a status update or a short brief, done within a time limit.
3. Run a role-play for customer-facing roles: a sales objection, an upset customer or a negotiation, with someone who will work with the hire.
4. Score against the bands above, not against an accent: the question is whether they can do the work at the level the role needs.
5. Use a standardized test only as a tiebreaker. EF offers free public versions of the EF SET, the same test behind the index.[6]

A role hiring guide adds interview questions and a scorecard for each role; English is one row of that scorecard, not the whole decision.