ESL Speaking Assessment Rubric: Free PDF and Examples
This ESL speaking assessment rubric uses four dimensions and four bands to describe what a learner does in a particular speaking task. Assess message and organisation, language resources, delivery and intelligibility, and interaction. Record a brief example before choosing a band, and use not observed when the task has not provided evidence.
Use it for formative pair work with teenagers or adults who can already make short suggestions and give simple reasons. The bands are original classroom descriptors, not CEFR levels, an examination scale or a validated placement test. Adapt the task and support to your learners; a score on one short task does not establish general proficiency.
Ready to use: 4-page printable pack, with all teaching content also available below. No account is needed.
Download the rubric and observer pack (PDF)Original classroom materials. You may print or adapt them for your own classes. Keep the source credit when sharing an adapted copy.
How to use the rubric fairly
- Share the goal and descriptors first. Choose one prompt and explain what a successful exchange includes. Let learners ask about unfamiliar task words.
- Give equal preparation and speaking opportunities. Allow two minutes to prepare keywords, then about four minutes for the pair exchange. Both learners must propose, respond and ask a question; a monologue alone cannot demonstrate interaction.
- Observe an individual, not just the pair. Write one short quote or observable action for each dimension. Record any sentence frames, extra preparation, repetition or teacher help used.
- Choose the best-supported band in each dimension. A learner may have an uneven profile. Do not give four identical scores just because you have an overall impression.
- Give one specific next step, then repeat with a comparable task. For example: ask a follow-up question that uses something your partner just said. Keep task demands and support similar when comparing attempts.
Focus on how understandable the message is, not how closely a learner imitates a particular accent. Do not score eye contact, personality, confidence, physical gesture or speed alone. Check that sound quality, unfamiliar task content or a lack of turn-taking opportunity is not being mistaken for a language difficulty. Provide agreed access arrangements and record them.
Four dimensions, four bands
Message and organisation
- 4 - Secure for this task
- Completes the task, gives relevant reasons and connects ideas so the listener can follow the message.
- 3 - Mostly independent
- Communicates the main answer in a mostly clear order; a reason or connection sometimes needs clarification.
- 2 - Developing with support
- Communicates part of the answer; needs a prompt to add a reason or put ideas in order.
- 1 - Needs substantial support
- Gives little relevant information; the main message remains unclear even after a supportive prompt.
Language resources
- 4 - Secure for this task
- Uses varied, relevant words and sentence patterns. Errors do not obscure the intended message.
- 3 - Mostly independent
- Uses enough words and basic sentence patterns for the task. Some errors occur, but meaning is usually clear.
- 2 - Developing with support
- Relies on a small set of words and patterns. Errors sometimes obscure meaning; asks for or needs language help.
- 1 - Needs substantial support
- Frequently cannot express the intended meaning with available words or sentence patterns, even with support.
Delivery and intelligibility
- 4 - Secure for this task
- Is understood throughout. Pace, pauses and emphasis help the listener; repairs a difficulty without losing the message.
- 3 - Mostly independent
- Is usually understood. Occasional pauses or repetition are needed, but the speaker can continue the message.
- 2 - Developing with support
- Frequent pauses or repetition interrupt understanding; the listener can still recover some key information.
- 1 - Needs substantial support
- Much of the message remains difficult to understand after repetition; sustained listener support is needed.
Interaction
- 4 - Secure for this task
- Initiates and responds, asks a relevant follow-up question, and helps the pair reach a shared decision.
- 3 - Mostly independent
- Responds relevantly and takes turns. Asks a question or checks a point with little prompting.
- 2 - Developing with support
- Responds to direct questions but needs prompting to take a turn, ask a question or check meaning.
- 1 - Needs substantial support
- Rarely responds to or builds on a partner's contribution, even when given an invitation and time.
Not observed (N/O): no adequate evidence was available for that dimension. Write why, such as “partner took all turns” or “recording failed”, then collect another sample. N/O is not zero and is not the same as band 1. Use band 1 only when a fair opportunity produced evidence that matches that descriptor.
Optional total: add the four scores only if all four dimensions were observed: 4-16 points. Keep the dimension profile and evidence alongside the total. If any dimension is N/O, leave the total blank. Do not turn the total into a CEFR level, percentage pass mark or automatic placement decision.
Three original speaking prompts
Use one prompt per observation. Allow keywords, not a full script, for the standard version. Offer sentence frames when needed and record that support. If assessing progress, do not silently compare a scripted first task with a much less supported second task.
1. Choose a class event
Your class has 40 minutes for an end-of-course event. Choose a quiz, a board-game session or a short film discussion. Tell your partner your choice and give two reasons. Ask which option they prefer. Agree on one plan and explain how everyone can take part.
Follow-up: If your partner prefers another option, ask: What would make that option work for everyone?
2. Solve a timetable problem
You and a partner need 30 minutes to finish a class project. You are both free on Tuesday at 16:00 and Thursday at 12:30. Tuesday has a quiet room but no internet. Thursday has internet but the room is shared with another group. Choose a time and place, explain one advantage and one drawback, and agree what to prepare first.
Follow-up: Ask one question about what your partner needs to finish the project.
3. Recommend a useful change
A learning centre can make one change: add a quiet study corner, lend headphones or start a conversation club. Explain which change you recommend and who it would help. Your partner proposes a different change. Ask a follow-up question, compare the two ideas and make a joint recommendation.
Follow-up: Explain what you would do if the centre could not provide your first choice.
Useful support: “I suggest ___ because ___.” “What do you think about ___?” “Do you mean ___?” “I see your point, but ___.” “Shall we agree on ___?” Give all learners access to the same support when making a like-for-like comparison.
A reusable observer sheet
Copy this structure into your notes or print page 3 of the PDF. Do not include names in a public or shared report; keep learner records in your usual secure school system.
Learner/code: __________ Date: __________
Task: __________ Preparation: __________
Support and access arrangements: ______________________________
| Dimension | Band or N/O | Words/actions observed |
|---|---|---|
| Message and organisation | ____ | ________________ |
| Language resources | ____ | ________________ |
| Delivery and intelligibility | ____ | ________________ |
| Interaction | ____ | ________________ |
Optional total (only four observed dimensions): ___ / 16
One successful action: ______________________________
One next step: _____________________________________
Next comparable task and date: _______________________
Two fictional scored examples
These are invented observation notes, not recordings, research results or evidence of classroom testing. The scores illustrate how to connect a judgement to evidence; delivery scores cannot be inferred from a written transcript alone.
Example 1: Maya chooses a class event
Conditions: two minutes of preparation; keyword notes; no sentence frames. Maya recommends a quiz because pairs can help each other and everyone can answer. She connects her reasons with “because” and “also”. The fictional observer notes that she is usually understood, with two pauses while searching for words. She answers her partner, but the teacher must remind her to ask a question.
- Message 4: a clear choice, two relevant reasons and an organised explanation.
- Language 3: enough basic language; she repeats “good” and uses “everyone can answers”, but the meaning remains clear.
- Delivery 3: the observer heard a generally understandable exchange with occasional pauses.
- Interaction 2: she responds, but needs prompting to ask a question.
Total: 12/16. Next step: ask “Which activity would you choose, and why?” without a reminder, then respond to the reason. A clear explanation has not hidden the need to work on interaction.
Example 2: Leo solves a timetable problem
Conditions: the same preparation time and keyword-note policy. Leo chooses Thursday and gives internet access as a reason, but needs a question to explain how the shared room affects the plan. The fictional observer reports occasional pauses without loss of the main message. Leo asks what his partner needs online, checks an unclear answer and suggests that they prepare their files first.
- Message 2: the drawback needs a prompt before the plan is sufficiently explained.
- Language 3: suitable basic vocabulary and sentence patterns communicate the plan.
- Delivery 3: the observer notes that the exchange is usually understandable.
- Interaction 4: he asks a relevant follow-up, checks meaning and helps reach agreement.
Total: 12/16. Next step: use “One drawback is ___, so we could ___” when presenting an initial recommendation. The same total as Maya represents a different teaching need.
Turn the observation into the next lesson
Choose a short follow-up that targets the weakest evidenced dimension: organise a recommendation into choice/reason/drawback; rehearse two useful sentence patterns; pause between ideas and stress key words; or ask a question that builds on a partner’s answer. Re-observe the action, rather than assigning more practice with no feedback goal.
For a complete shared task with optional scaffolds, use the 45-minute mixed-ability trip-planning lesson. To distinguish activity completion from evidence of learning, see ESL progress tracking for teachers. The downloadable rubric is a teacher observation tool; it does not claim that the platform automatically scores this rubric.
Source notes and authorship
The Council of Europe’s CEFR descriptor guidance and 2020 Companion Volume distinguish aspects of spoken performance, including interaction and intelligibility. They informed the choice to observe separate dimensions here. The four bands, prompts and examples on this page are original materials by The Kingdom of English, not official Council of Europe descriptors or an endorsed assessment instrument. Sources checked October 3, 2026.