Some people get a high TOEIC score but cannot speak English well. Why? The TOEIC Listening and Reading test measures listening and reading. It does not measure speaking directly.

A 2008 study found that people with higher scores said they could do more listening and reading tasks. To know your speaking level, a speaking test works best. But listening and reading are still the base of speaking, so practice speaking, too.

Many people wonder why someone with a TOEIC score of 900 may still struggle to speak English. The answer lies in what the test measures. In a 2008 study, Powers and his colleagues asked TOEIC Listening and Reading test takers in Japan and Korea how well they could do everyday English tasks. People with higher scores consistently reported being able to do more listening and reading tasks.

A later study of about 2,300 test takers found that speaking ability was best predicted by a speaking test score. Adding listening and other scores made the prediction even more accurate. So a high score is not false, but it does not directly show speaking ability. Learners who want to speak should practice speaking separately.

The familiar complaint that people with TOEIC scores of 900 cannot hold a conversation often reflects a misunderstanding of what the Listening and Reading test measures. Powers, Kim and Weng asked test takers in Japan and Korea to rate how well they could perform everyday English tasks. Higher scores were consistently associated with more "can do" responses for listening and reading, and the two sections appeared to tap partly distinct abilities.

In'nami and Koizumi confirmed that listening and reading form separate but strongly related factors. Speaking is another matter. Analyzing self-assessments from about 2,300 test takers, Powers and Powers found that speaking ability was predicted primarily by TOEIC Speaking scores, although adding scores from other sections, including listening, increased predictive accuracy. A high score is not misleading; it simply does not measure speaking directly.

Two limitations deserve mention. The criterion in these studies was usually self-assessment, which may be affected by confidence, and much of the research was conducted by ETS, the organization that develops the test, so independent verification is valuable. For learners, the practical implication is to treat the score as a foundation and to train speaking separately.