TOEFL adaptive testing: can your score be low after an upper module?

How does the adaptive TOEFL choose the second module?

ETS says that performance on the first part of Reading and Listening determines the content of the second part (ETS "Updated TOEFL iBT Test Overview", page 3, hereafter the Test Overview; no publication date, checked 2026-10-30). The Test Overview describes both sections as multistage adaptive (the questions in the second part depend on your answers in the first part), with the aim of matching task difficulty to the test taker's ability level.

The Technical Manual says the first module (Router) is built from medium-difficulty tasks (CEFR B1 or B2), and that doing well on the first module makes the second module harder (page 28). Each test taker takes one of two paths: "Router + lower module" or "Router + upper module" (same page). ETS "TOEFL iBT Test: 2026 Update Test Blueprint and Specifications Document" (hereafter the Specifications Document) labels the first module Router and the second modules Lower Module and Upper Module (Specifications Document, page 2 table; no publication date, checked 2026-10-30). In this article, "upper module" means the Upper Module. Writing and Speaking are linear (every test taker who receives the same test form answers the same tasks) (Specifications Document, page 3, note 5).

The table comes from the page 2 table of the Specifications Document (checked 2026-10-30). The Specifications Document says to count total time as the first module plus either the lower or the upper module (page 3, note 5). ETS "TOEFL iBT Test Content" also says that because the test adapts, test time and questions can vary (no publication date, checked 2026-10-30).

How are Reading and Listening scored?

ETS says Reading and Listening questions are scored 1 point if correct and 0 points if wrong, and that the sum of points in a section (the raw score) is converted to a scaled score through equating (Technical Manual, page 34, checked 2026-10-30). The Specifications Document also lists both sections as machine scored (scored by fixed rules, such as matching the answer key) (page 2 table and page 3, note 4).

Here is what Technical Manual page 34 says about equating.

• A separate conversion table from raw scores to scaled scores is made for each test form.
• Equating accounts for the differences in difficulty created by multistage adaptation.
• The same reported score means the same language ability, whichever second module the test taker received.
• A reported score is not the same as the number or percentage of questions answered correctly.

We read these sentences to mean that getting the upper module does not by itself decide your section band, and that the number of questions you answered correctly in both modules also counts.

Some questions are not scored. The Technical Manual says Reading modules can contain 15 unscored questions and Listening modules may include 12, and that the first module can also contain questions being tried out (page 28). Each path has 35 scored questions in both sections (page 29, Table 1). ETS does not say which questions are unscored.

The score scale is the 1 to 6 band. The Test Overview says the four section scores and the total score are reported on the 1 to 6 band in 0.5 steps, rounded to the nearest whole number or 0.5 (page 21). The total score is the average of the four section bands (same page). In ETS's examples, a four-section average of 5.125 gives a total score of 5, and an average of 5.25 gives a total score of 5.5 (1 to 6 band; ETS "TOEFL iBT Test Content" and "Understanding TOEFL Scores", checked 2026-10-30). The rounding and the comparable 0 to 120 total score ranges are shown in a table in the Unlock Prep (unlockprep.com) guide "How do TOEFL 1 to 6 scores compare with 0 to 120 scores?".

How do people misread their score after an upper module?

The misreading behind this question is the belief that getting the upper module means a high score. The three readings below conflict with sentences in ETS documents. They are examples we wrote for illustration.

Example 1

• Wrong reading: "On the TOEFL, every question you answer correctly makes the next question harder."
• Right reading: "In Reading and Listening, performance on the first module selects one second module; the test does not adapt question by question."
• Reason: The Technical Manual says adaptation happens at the section level, not at the level of individual questions (page 28). The Specifications Document also describes both sections as two-stage adaptive (page 3, note 5).

Example 2

• Wrong reading: "If you got the upper module, your Reading band (1 to 6) will be high no matter how many questions you answered correctly."
• Right reading: "The Reading band reflects both the number of questions answered correctly in the two modules and the difficulty of the modules received."
• Reason: Technical Manual page 28 says both section scores take into account the number of correct answers in the two modules and the difficulty of the modules on the path. Page 34 says the same reported score means the same ability, whichever second module the test taker received.

Example 3

• Wrong reading: "I got the upper module in Reading and Listening, so my total score (1 to 6 band) cannot be low."
• Right reading: "The total score is the average of the four section bands, so even with high Reading and Listening bands, the total score can be low depending on the Writing and Speaking bands."
• Reason: The total score is the average of the four section bands, rounded to the nearest 0.5 (Test Overview, page 21). With section bands of Reading 6, Listening 6, Writing 3 and Speaking 3, the average is 4.5 and the total score is 4.5 (1 to 6 band; our calculation). Even with Reading and Listening at 6, the top of the scale, the total score stays at 4.5.

What has ETS said, and what has it not said?

ETS has explained how the second module is chosen and how scoring reflects module difficulty, but it does not publish the conversion tables used in the calculation or the cut score that leads to each second module.

What ETS has said (checked 2026-10-30)

• Reading and Listening are two-stage adaptive, and doing well on the first module makes the second module harder. Writing and Speaking are linear (Test Overview, page 3; Technical Manual, page 28; Specifications Document, page 3, note 5).
• There are two second modules, a lower module and an upper module, and in Listening their estimated times differ: 7 minutes and 11 minutes (Specifications Document, page 2 table).
• Both section scores are set by equating, reflecting the number of correct answers in the two modules and the module difficulty, and the same reported score means the same ability whichever second module the test taker received (Technical Manual, pages 28 and 34).
• Each question is scored 1 or 0, each path has 35 scored questions, and Reading modules can contain 15 unscored questions while Listening modules may include 12 (Technical Manual, pages 28, 29 and 34).
• Section scores and the total score are rounded to 0.5 steps on the 1 to 6 band, and the total score is the average of the four sections (Test Overview, page 21).

What we did not find in ETS documents (Test Overview pages 1 to 28, Specifications Document pages 1 to 18, Technical Manual pages 28, 29, 34 and 41 plus a full word search, the Information Bulletin, and the ets.org Test Content, Understanding TOEFL Scores, Reading and Listening pages, 2026-10-30)

• How many first-module questions you must answer correctly to reach the upper module (the cut score)
• The numbers in the conversion tables from raw scores to the 1 to 6 band (made separately for each test form and not published; Technical Manual, page 34)
• Whether getting the upper module puts a floor under your band, or getting the lower module puts a ceiling on it
• Which questions are unscored
• Whether test takers are told which module they received
• How much unofficial scores can differ from official scores (ETS says only that changes are rare; see section 5 below)

Where sources differ: the Reading and Listening raw score ranges are not the same across documents.

Of the three documents, the Specifications Document and the Technical Manual give 35 for Reading and Listening, and only the Test Overview gives 0 to 30. The 35 scored questions per path in Technical Manual page 29, Table 1 match the figure of 35. This article does not calculate scores from these numbers.

Our view: do not guess your section bands from which second module you got; read them from the four section bands on your score report.

Where do you check if your score is lower than expected?

First put the four section bands (1 to 6) on your score report side by side, and see which section pulled the total score down.

• The Reading and Listening scores on the screen at the end of the test are unofficial scores (ETS 2026-27 "TOEFL iBT Information Bulletin", page 11, hereafter the Information Bulletin; checked 2026-10-30). ETS "Understanding TOEFL Scores" says it is rare, but possible, for these scores to change after quality checks (checked 2026-10-30).
• Official scores appear in your ETS account 3 days after the test date (ETS "Understanding TOEFL Scores" and Test Overview, page 3, checked 2026-10-30).
• Score reviews are available only for Speaking and Writing. The Information Bulletin says you can request a review of these two sections for a fee in your ETS account within 30 days after the test date, that you cannot request one if you have already asked for your scores to be sent to institutions, and that the resulting score can go up or down (page 12). ETS "Understanding TOEFL Scores" says Reading and Listening scores cannot be reviewed or disputed (checked 2026-10-30).
• Check your target score in the notices of the institutions you are applying to. ETS says it has no passing or failing score and that each institution sets its own requirements (ETS "Understanding TOEFL Scores", checked 2026-10-30). How to read institution requirements is covered in "What TOEFL score do I need?".
• How to read mock test scores against real scores is covered in "How closely do TOEFL mock test scores match real scores?".