How TOEFL Listening Is Scored in 2026

How TOEFL Listening is scored in 2026 is a count of questions you got right, then a section result on the official 1–6 scale. In practice on this site, each question is right or wrong and the indicative 1.0–6.0 band is mapped from that count. There is no partial credit for a reply that was polite, a detail that was mentioned, or an inference that was interesting. The band orients the next drill. It is not an ETS score.

In this guide

1. What earns the point
2. Points by task
3. Modules add items, not a bonus
4. How to read the report
5. Comparisons that mislead
6. Try this yourself
7. Frequently asked questions

What earns the point

A Listening question earns one point when the option you select is the keyed option. The audio is the evidence. A choice that repeats a word you remember is not closer to correct than a choice that misses the topic entirely. Both are 0 if they are not the key.

You do not get a speaking-style criterion score for “good listening.” Fluency, intelligibility, and organization belong to Speaking. Listening does not grade how you would have replied aloud. On Choose a Response you select a written reply. On the other tasks you select answers about what you heard. Confusing those reports is how people practice the wrong repair.

The official scale is 1–6 in half steps. ETS does not hand out a public chart that turns a two-question conversation into an official band. Anyone who offers that chart is guessing. Use your fraction.

Points by task

Choose a Response is one point per remark. The point is the function: the reply answers the speaker. The response guide shows a remark where the correct line answers a surprise, and the traps echo “library.”

A conversation practice item on this site is two points, one per question. Typically one asks what the problem is and one asks what someone tells the other person to do. Missing only the instruction is still a missed point. It is not “basically right.”

An announcement practice item is also two points. One may ask the reason or the new time. One may ask what you should do. A usual campus policy that the speaker did not say does not score.

An academic-talk practice item is three points. Main idea, a relationship in the talk, and the purpose of an example are the usual jobs. A detail that is true and is not the example’s job scores 0 on the purpose question. The talk guide works one example so the purpose point is visible.

Task in solo practice	Points available	Zero sounds like
Choose a Response	1 per remark	“It repeated the noun”
Conversation	2	“I got the problem and guessed the ending”
Announcement	2	“That’s what announcements usually do”
Academic talk	3	“That fact was in the talk”

The full section has many more items than a solo task. A perfect two-question drill is not a Listening section score. The section guide keeps the public item count and the practice clocks apart.

Modules add items, not a bonus

Module 2 is easier or harder because of Module 1. Those items join the count. They do not multiply Module 1, and reaching the harder path is not an extra point. If you miss the same announcement pattern in both modules, the route did its job: it showed the pattern is stable. The repair is an announcement task for 1 credit, not another 3-credit section to chase the hard path.

Read the module split in the report before you read the band. A band can sit on the same half step while one module was clean and the other was not. The half step is a rounded summary. The split is the decision.

How to read the report

Write the fraction first: correct over total, for the task you took. Then open only the misses. For each, name the job that failed: function of the reply, last instruction, stated change, or purpose of the example. If the name is a topic (“the library,” “psychology”), rename it. Topics do not transfer to the next clip.

The rationale points at what the correct option used. If you cannot say that in one line, you have not finished the report, and another credit will spend the same miss. TOEFL Listening practice should be opened after that line exists, not instead of it.

Comparisons that mislead

Do not average a Listening practice band with a Reading practice band and call it a total. Do not compare a three-question talk with a ten-blank Complete the Words set. Do not treat a Speaking 0–5 item as if it were a Listening question. Reading uses the same right-or-wrong idea and still is not the same task.

Do not replay the clip in a separate tab, answer, and then trust the band. The band describes the attempt the product recorded. A second private hearing describes a different attempt. If you need the transcript to understand a word, do that after the report, as study, and do not mix it into the count.

Same band, different decision

Two reports can both say 4.0 and demand opposite next steps. In one, you missed three Choose a Response items and nothing else. The next credit is that task, and the repair is function versus echo. In the other, response items were clean and both announcement questions missed the clock. The next credit is an announcement, and the repair is a labeled time. Averaging those sessions into “Listening is a 4” erases the only fact that chooses the URL. The band is allowed to match. The pile is not allowed to disappear.

Write the pile on the report in the margin if you print it, or in the first line of a note if you do not. Include the fraction beside it. A week later you will not remember whether 4.0 was the echo week or the clock week. You will remember a number that cannot tell those weeks apart. The 14-day plan is built on piles for that reason. If you skip the pile, the calendar still spends credits and you will feel busy. Busy is not a count that moved.

Also separate a solo task’s band from a section’s band even when the digits match. A 4.0 on eight response items is a tight claim about one job. A 4.0 on a two-module mix is a blend. Do not celebrate the blend until you know which job is inside it. The harder module does not explain the blend by itself. Open the module split, name the task, then decide.

Try this yourself

Take one Choose a Response set. Before you look at the band, write how many remarks you think you answered. Then open the count. If the band is the only thing you remember an hour later, retell the count from memory before you start anything else.

Start Choose a Response. Use the full Listening section when you want both modules in that count, which is a 3-credit start.

Frequently asked questions

Is there partial credit in Listening?

Not on these practice items. The keyed option scores 1. Every other choice scores 0, including choices that contain a real detail from the audio.

Does the harder module raise my score automatically?

No. It adds items. The indicative band follows how many were correct. Routing is not a bonus and not a penalty you can see as a separate subtraction.

Why did my band stay the same when I got one more question right?

The displayed figure rounds to the nearest 0.5. Small count changes can sit inside one step. Read the fraction. The step is a label.

Can this match my official Listening score?

No. Official scoring is ETS’s. This report is an indicative practice figure from the items you answered here.

Next steps

Put the fraction into the 14-day Listening plan on today’s date. The next credit should be the task that owned the misses, not the task you like hearing. If the fraction and the band disagree about whether you improved, keep the fraction and let the half-step catch up on a later set.