IELTS Speaking Part 2 Scoring: The Four Criteria Explained

The IELTS Speaking Part 2 long turn places candidates alone with a topic card for 120 seconds of uninterrupted speech. Unlike timed interviews in other assessments, this section tests the ability to sustain a well-structured, relevant, and linguistically varied monologue. Understanding how examiners apply the four official marking criteria — Fluency and Coherence, Lexical Resource, Grammatical Range and Accuracy, and Pronunciation — transforms preparation from guesswork into targeted skill-building. This article walks through each criterion in detail, contrasting the behaviours that characterise different band scores and offering concrete guidance for candidates aiming to move from Band 6 to Band 7 or beyond.

How the IELTS Speaking Part 2 marking framework operates

The IELTS Speaking test is marked by trained examiners who hold official certification from the British Council, IDP, or Cambridge Assessment English. Each examiner evaluates all three parts of the speaking test — including Part 2 — against four independent criteria simultaneously. No single criterion carries disproportionate weight; a strong lexical score cannot compensate for persistent grammatical errors, nor can flawless pronunciation rescue a candidate who struggles to maintain coherence.

For Part 2 specifically, the examiner is listening for the candidate's ability to address the four bullet points on the cue card while demonstrating range and control across all four criteria. The bullet points exist precisely to prevent candidates from delivering a generic, content-light speech; they function as a structural scaffold and a content relevance check. The official band descriptors are public documents, and studying them directly — rather than relying solely on third-party summaries — gives candidates the most accurate picture of what is required.

Fluency and Coherence in the long turn

Fluency and Coherence is often the criterion candidates find most ambiguous. It does not simply mean speaking quickly. The official band descriptors describe it as the ability to speak at length without noticeable effort, to produce spontaneous and seamless discourse, and to use a wide range of cohesive devices appropriately. In Part 2, the examiner is specifically assessing whether the candidate can maintain a continuous monologue for the full two minutes without freezing, repeating, or abandoning the topic.

At Band 6, candidates typically display adequate fluency: they can speak at length with acceptable flow, though they may self-correct or hesitate occasionally. Their coherence is evident through logical ordering of ideas, but cohesive devices may be somewhat predictable — words and phrases such as first of all, also, and in conclusion appear frequently. The discourse is generally easy to follow.

At Band 7, the descriptor calls for few hesitations and rarely re-phrasing or searching for language. Candidates should demonstrate that they can talk at length without undue pausing, and that their use of cohesive devices is flexible and accurate rather than formulaic. The transition from Band 6 to Band 7 in this criterion often hinges on the perception of spontaneity: a Band 7 speaker sounds as though they are sharing a story rather than reciting a memorised script.

A common misconception is that pausing is always penalised. Brief, natural pauses — for example, a moment to gather thought before a significant detail — are entirely acceptable and do not lower scores. What examiners penalise is prolonged silence, repeated restarts, and topic abandonment. Candidates should practise sustaining speech by training themselves to narrate in real time, describing events as though they are occurring in the present, which naturally reduces hesitation.

Lexical Resource: vocabulary range and precision in Part 2

Lexical Resource assesses the breadth and accuracy of a candidate's vocabulary, as well as their ability to deploy words in context-appropriate ways. Part 2, with its requirement to describe a person, place, object, event, or activity, demands a substantial descriptive vocabulary — words for physical appearance, personality traits, emotional states, spatial relationships, and temporal sequencing. Candidates who rely on a narrow lexical set will find it difficult to sustain two minutes of speech without repetition.

The Band 6 descriptor calls for an adequate range of vocabulary for accurate communication on familiar topics, with occasional inappropriate word choices and some non-native phrasing. In practice, this means that a Band 6 candidate can convey the core meaning of their ideas but may resort to general words where more precise ones would serve better. For example, saying "a very nice place" for every location described, rather than varying between picturesque, bustling, serene, or architecturally striking.

Band 7 requires a wide enough range to allow some flexibility and precision, with occasional slight inaccuracies in word choice or collocation. The distinction from Band 6 lies in the ability to use less common vocabulary in contextually appropriate ways. A Band 7 candidate might describe a family member using terms such as stoic, affable, or unassuming, and use precise collocations such as deeply rooted tradition or profoundly formative experience.

Candidates should deliberately expand their topic-specific vocabulary for each of the four Part 2 question families — describing a person, place, object, activity, and event. Vocabulary building should not focus solely on individual words but on lexical chunks: ready-made phrases such as on the spur of the moment, the sort of person who, or if memory serves me correctly add naturalness and reduce processing time during the test. Using a dictionary alongside example sentences from corpus sources (such as the Cambridge English Corpus) helps ensure that new words are learned in their typical collocational patterns.

Grammatical Range and Accuracy: the structural backbone of sustained speech

Grammatical Range and Accuracy evaluates the diversity of sentence structures a candidate can produce and how consistently those structures are error-free. In Part 2, candidates are expected to narrate past events — which requires confident use of the past tense — and to describe habitual actions, ongoing situations, and hypothetical scenarios. This variety of temporal reference demands control over multiple grammatical forms, not just the simple past.

At Band 6, the descriptor describes a mix of simple and complex sentence forms, with a fair degree of accuracy but errors that do not impede communication. Grammatical errors at this level are typically tense inconsistency, article omissions, or subject-verb agreement slips that the listener can still parse without difficulty. The candidate uses some complex structures — subordinate clauses, passive voice, relative clauses — but these may be structurally flawed or appear only intermittently.

At Band 7, candidates must demonstrate frequent error-free sentences and a variety of complex structures used with flexibility and accuracy. This does not mean zero errors; the band descriptor explicitly allows for occasional errors. What characterises Band 7 grammatical performance is the ability to produce a range of structures — conditional sentences, embedded clauses, cleft sentences, participial constructions — without those structures breaking down mid-utterance.

A specific challenge in Part 2 is maintaining past tense accuracy throughout a sustained narrative. Candidates frequently shift tenses mid-sentence or revert to present tense when describing past events, particularly under the cognitive load of the two-minute speaking requirement. Practising timed past-tense narrations on a range of topics — from a childhood memory to a recent holiday — builds the automaticity needed to sustain tense consistency under exam conditions. Working with a tutor or language partner who can flag tense drift during practice sessions is particularly effective for this criterion.

Pronunciation in the context of a two-minute monologue

Pronunciation is assessed at the level of individual sounds, word stress, sentence stress, and intonation patterns. The IELTS pronunciation marking criterion is not a test of accent; candidates are not expected to sound like native speakers of British or American English. The descriptors focus on the degree to which a candidate's pronunciation is easy to understand and on the effectiveness of their use of intonation, stress, and rhythm to convey meaning.

Band 6 requires that pronunciation be largely intelligible throughout, though there may be noticeable concentration of errors in particular sounds, word stress, or intonation patterns. A Band 6 candidate can generally be understood, but the examiner may need to listen more carefully, and some individual sounds may cause temporary confusion.

Band 7 demands intuitive pronunciation, meaning the candidate's speech patterns are largely automatic and do not require conscious effort from the listener to decode. This includes appropriate use of thought groups — pausing at natural clause boundaries rather than mid-thought — and correct placement of stress on content words within sentences. A Band 7 speaker signals new or important information through stress and uses intonation to signal the structure of their discourse — for instance, rising intonation to introduce a list, falling intonation to conclude a point.

The most effective pronunciation improvement strategy for Part 2 is shadowing practice: listening to native or near-native speakers delivering short talks or descriptions, then immediately reproducing what was heard with attention to rhythm, stress, and intonation. Recording oneself describing various Part 2 topics and comparing the playback to native-speaker models helps identify persistent problem sounds and unnatural stress patterns.

Common marking criterion pitfalls and how to avoid them

Several recurring patterns consistently push candidates below the band score they might otherwise achieve in Part 2. Identifying and correcting these habits during preparation is far more efficient than discovering them on test day.

The first pitfall is over-relying on memorised phrases rather than genuine fluency. Some preparation courses encourage candidates to memorise generic openings such as "The person I would like to talk about is..." and closings such as "That is the person I wanted to describe." While these phrases are not wrong, heavy reliance on them signals to the examiner that the candidate may be using template-based responses. Band 7 and above require evidence of genuine, spontaneous language production. The solution is to practise with diverse topics so that language is generated naturally rather than retrieved from a script.

The second pitfall is ignoring one or more of the four cue card bullet points. Each bullet point corresponds to roughly 30 seconds of content in a well-balanced Part 2 response. Candidates who focus exclusively on the first bullet and rush through the remaining three — or who skip a bullet entirely — are judged to be providing incomplete responses. The examiner checks whether all four points have been addressed, and incomplete coverage inevitably caps the available score.

The third pitfall is using oversimplified narrative structures. Some candidates tell their Part 2 stories as simple linear sequences — "First I did this, then I did that, then it ended" — without sub-clauses, without parenthetical details, and without evaluative commentary. This produces grammatically simple (though often accurate) speech that lacks the structural complexity the Band 7 grammatical criterion requires. Introducing embedded clauses ("The meal, which we had been looking forward to for weeks, turned out to be disappointing"), relative clauses ("My grandmother, who was born in 1948, still lives in the house where she grew up"), and adverbial clauses of cause and effect ("I decided to change universities because the course did not match my expectations") immediately enriches the grammatical texture of the response.

Finally, candidates should avoid excessive self-correction. A small amount of self-correction demonstrates monitoring ability and is appropriate at Band 7 and above. However, repeatedly stopping mid-sentence to restart — "I went to — no, I visited — the place where — actually, let me start again" — disrupts fluency and coherence to the point where it lowers the score.

Band score comparison across the four criteria in Part 2

The following table illustrates the key distinctions between Band 6 and Band 7 across all four marking criteria as they apply to the Part 2 long turn context.

Criterion	Band 6 characteristics	Band 7 characteristics
Fluency and Coherence	Speaks at length with acceptable flow; occasional hesitations and self-correction; uses simple cohesive devices	Few hesitations; rarely re-phrases or searches for language; flexible and appropriate use of cohesive devices
Lexical Resource	Adequate range for familiar topics; occasional inappropriate word choices; some repetition	Wide enough range for flexible and precise expression; occasional slight inaccuracies in word choice or collocation
Grammatical Range and Accuracy	Mix of simple and complex forms; fair accuracy with errors that do not impede communication	Frequent error-free sentences; variety of complex structures used with flexibility and good control
Pronunciation	Largely intelligible; noticeable concentration of errors in specific sounds or stress patterns	Intuitive and mostly automatic; effective use of stress, intonation, and thought groups

Integrating criterion awareness into Part 2 preparation

Understanding the four marking criteria should not become a distraction during the test itself. Candidates who attempt to mentally monitor each criterion while speaking are likely to produce stiff, unnatural speech. Instead, criterion awareness should inform the preparation phase, where deliberate practice targets specific weaknesses identified through self-assessment or feedback from a qualified teacher.

A structured preparation approach begins with a diagnostic practice session: deliver a timed Part 2 response on an unfamiliar topic without interruption, recording both audio and a self-assessment against the four criteria. Compare the recording against the official band descriptors to identify priority areas. For a candidate whose grammatical range is the primary constraint, the preparation focus shifts to embedding clauses and expanding tense usage in timed narrations. For a candidate whose lexical resource is the constraint, the focus shifts to targeted vocabulary expansion within the five Part 2 topic categories.

Practice should always be timed — two minutes, without stopping — and should cover all four cue card bullet points in every session. Using past paper topics is essential, as the distribution of topics in the official Cambridge IELTS practice tests reflects the frequency patterns of real test administrations. Candidates should work through at least 20 to 30 different topics across their preparation period to build confidence with unfamiliar content.

Next steps for targeted preparation

The IELTS Speaking Part 2 long turn rewards candidates who understand not just what topics might appear, but how the four marking criteria translate into observable speaking behaviours. Targeted preparation — informed by the official band descriptors and focused on specific weakness areas — is considerably more effective than general conversation practice alone. Whether the priority is building grammatical complexity, expanding topic-specific vocabulary, developing more intuitive pronunciation, or sustaining fluency without scripted dependencies, each skill responds to deliberate, criterion-referenced practice.

Candidates seeking a structured starting point for diagnostic assessment and personalised study planning can explore TestPrep's complimentary initial consultation to identify which of the four criteria warrants the most focused attention during the preparation period ahead.

Frequently asked questions

Can the examiner fail me in Part 2 for not covering all four bullet points on the cue card?

The cue card provides four bullet points, and while the examiner will not stop you for omitting one, the assessment criteria reward complete coverage of the task. Addressing all four bullet points allows the examiner to evaluate your full range of language ability. Incomplete coverage typically caps the available score, particularly for Fluency and Coherence and Task Response, and signals a lower level of communicative competence.

Is it better to use more complex grammar even if it risks more errors, or to stay safe with simple sentences?

The band descriptors explicitly reward grammatical variety at Band 7 and above. A candidate who uses only simple sentences is unlikely to score above Band 6 regardless of accuracy. The appropriate strategy is to produce complex structures you can control — sentences you have practised and can deploy accurately in real time. Occasional errors are permitted at Band 7; the descriptor refers to frequent error-free sentences, not perfect sentences.

How does Part 2 connect to the marking of the overall speaking test score?

The four criteria are applied to the speaking test as a whole, not independently to each part. Part 2 contributes a significant portion of the evidence for each criterion, but the examiner also considers Part 1 (the introduction and interview) and Part 3 (the two-way discussion) when arriving at a final band score for each criterion. Strong performance in Part 2 can compensate somewhat for weaker performance in Part 1, and vice versa, but consistency across all three parts is the most reliable path to a high overall score.

Does a non-native accent reduce my pronunciation score automatically?

No. The pronunciation criterion is not a test of accent. Examiners are trained to assess clarity, intelligibility, and the effective use of stress, intonation, and rhythm — not whether the candidate sounds like a British or American speaker. Candidates are assessed on how effectively they convey meaning through pronunciation, not on the regional variety of English they use. Consistent errors in specific individual sounds, however, can affect the overall intelligibility rating.

Should I memorise content for Part 2 topics in advance?

Memorising specific answers to anticipated topics is not recommended and can be counterproductive. The examiner is trained to identify memorised responses, which are typically delivered with a flat, non-interactive tone and may sound formulaic. More importantly, memorised content restricts flexibility: if the exact topic one has memorised does not appear, the candidate is left without a genuine fallback strategy. Instead, candidates should develop a repertoire of <em>language skills</em> — narrative structures, descriptive vocabulary, cohesive devices — that can be applied to any topic within the five Part 2 question families.

How IELTS Examiners Score Speaking Part 2 Long Turn