At B1, reading questions ask what a text says. At B2, they ask what the writer is doing.
Why did the author put that concession in the fifth paragraph? What is the function of that statistic? Is the tone hostile or measured? Where does the writer stand, given that half the piece argues against itself?
You cannot answer those by finding a sentence. You have to see the architecture of the argument. That is why B2 reading defeats candidates whose vocabulary is otherwise fine.
This sample paper is built around exactly that skill.
B2 reading and B2 writing lean on the same understanding of argument. DELF B2 written papers ask you to build the structure you are about to learn to read.
How Is the DELF B2 Reading Paper Structured?
Three exercises across around five documents, worth 25 marks in one hour. Texts run to roughly 500 words and come from the press, reports, and opinion writing on social topics. All answers are multiple choice.
The documents fall into two families. Informative texts present a situation or a set of findings. Argumentative texts take a position and defend it, usually while acknowledging the other side.
The second family is where the marks are decided. Informative texts test whether you caught the information. Argumentative texts test whether you followed the reasoning.
Official past papers are available for download from the B2 sample subjects page published by France Éducation International. The B2-level description from the Swiss DELF/DALF network sets out what the diploma certifies.
Reading sits in the middle of the longest collective session of any DELF level, which runs for two and a half hours. Pass is 50 out of 100, with a floor of 5 out of 25 here.
What Does B2 Ask That B1 Does Not?
Five things: the author’s position, the function of each paragraph, the purpose of an example or statistic, tone, and the difference between a view the writer holds and one they merely report.
That last distinction causes more errors than anything else. B2 texts often spend two paragraphs on a case that the author will then take apart. A candidate reading for content finds those paragraphs and answers from them.
Reformulation runs underneath all of it. The correct option almost never repeats the text’s wording, and wrong options frequently do.
This page is the practice paper. For the text families, question types, and the full preparation method, see the DELF B2 reading guide.
The Sample Paper
La gratuité des transports, fausse bonne idée ?
Depuis que Dunkerque a rendu son réseau de bus entièrement gratuit en 2018, l’idée fait son chemin. Une trentaine de villes françaises l’ont adoptée, et l’argument séduit : si le transport public ne coûte rien, chacun l’utilise, et la voiture recule.
Les résultats dunkerquois semblent d’ailleurs encourageants. La fréquentation a bondi de plus de soixante pour cent en deux ans, et une partie des nouveaux usagers déclare avoir laissé sa voiture au garage. Difficile de ne pas y voir un succès.
Pourtant, un examen plus attentif invite à la prudence. D’abord, la gratuité ne crée pas de bus. Elle attire des voyageurs sur un réseau dont la capacité, elle, reste inchangée. Dunkerque a dû, en parallèle, augmenter le nombre de lignes et la fréquence des passages. Ce sont ces investissements, autant que la gratuité, qui expliquent le report observé.
Ensuite, la question du financement reste entière. Les recettes des billets représentaient environ dix pour cent du budget du réseau dunkerquois : une part faible, donc supprimable sans dommage. Dans une métropole comme Lyon ou Paris, cette proportion dépasse largement les trente pour cent. Y renoncer supposerait soit une hausse d’impôts, soit une baisse de l’offre. Or c’est précisément l’offre qui détermine si les automobilistes changent d’habitude.
Enfin, plusieurs études signalent un effet inattendu. Une partie des nouveaux usagers ne vient pas de la voiture mais du vélo ou de la marche. Autrement dit, la gratuité peut détourner des modes de déplacement déjà vertueux, sans réduire la circulation automobile autant qu’espéré.
Faut-il pour autant écarter l’idée ? Ce serait excessif. Dans les villes moyennes, où les recettes billettiques sont faibles et le réseau sous-utilisé, la gratuité présente un rapport coût-bénéfice réel. Elle constitue aussi une mesure sociale directe pour les ménages modestes, ce qui n’est pas rien.
Le débat gagnerait simplement à être posé autrement. La vraie question n’est pas « gratuit ou payant », mais « où placer l’argent public ». Un réseau payant, dense et fiable convainc davantage d’automobilistes qu’un réseau gratuit qui passe toutes les quarante minutes. Ceux qui présentent la gratuité comme la solution évacuent ce choix, et c’est là que le raisonnement pèche.
1. Selon l’auteur, les résultats de Dunkerque s’expliquent… a) uniquement par la gratuité b) par la gratuité et par des investissements dans l’offre c) par la hausse du prix des carburants
2. Que signifie « la gratuité ne crée pas de bus » ? a) les bus existants sont supprimés b) la mesure n’augmente pas la capacité du réseau c) les bus circulent moins souvent qu’avant
3. Pourquoi la gratuité serait-elle plus difficile à mettre en place à Lyon qu’à Dunkerque ? a) le réseau lyonnais est moins fréquenté b) les recettes des billets y représentent une part bien plus importante du budget c) les habitants y seraient hostiles à la mesure
4. L’effet inattendu signalé par l’auteur est que… a) certains nouveaux usagers renoncent au vélo ou à la marche b) les automobilistes circulent davantage qu’avant c) les bus restent vides aux heures creuses
5. Quelle est la position de l’auteur sur la gratuité ? a) il la rejette dans tous les cas b) il la juge pertinente dans certains contextes seulement c) il la soutient sans réserve
6. Dans l’avant-dernier paragraphe, l’auteur cherche surtout à… a) nuancer sa propre critique b) résumer les arguments de ses adversaires c) attaquer la mesure une nouvelle fois
7. Pour l’auteur, la question centrale du débat porte sur… a) le prix du ticket b) l’affectation de l’argent public c) le nombre de voitures en circulation
8. Le ton général du texte est… a) polémique et hostile b) critique mais mesuré c) franchement enthousiaste
Answer Key, and What Each Question Really Tests
1. b) par la gratuité et par des investissements dans l’offre. Paragraph three says “Ce sont ces investissements, autant que la gratuité, qui expliquent le report observé.” The phrase autant que is the whole answer. Option A repeats the framing of paragraphs one and two, which the author sets up only to take apart. Tests: telling a reported view from the author’s own.
2. b) La mesure n’augmente pas la capacité du réseau. The following sentence explains it: more travelers on a network whose capacity remains unchanged. Tests: interpreting a compressed rhetorical phrase from its context.
3. b) Les recettes des billets y représentent une part bien plus importante du budget. Ten percent against more than thirty. The text never says Lyon is harder in those words, so you have to draw the comparison. Tests: reasoning from two figures.
4. a) Certains nouveaux usagers renoncent au vélo ou à la marche. “Ne vient pas de la voiture, mais du vélo ou de la marche”, puis “détourner des modes de déplacement déjà vertueux.” Tests: reformulation across two sentences.
5. b) Il la juge pertinente dans certains contextes seulement. The sixth paragraph concedes a real case for medium-sized towns. Stop reading at paragraph five, and you pick option A. Tests: holding the whole argument in view.
6. a) Nuancer sa propre critique. This is the paragraph-function question, the B2 signature. The paragraph opens with a rhetorical question, answers “ce serait excessif”, then gives ground. Tests: seeing what a paragraph does, not what it says.
7. b) l’affectation de l’argent public. “La vraie question n’est pas de savoir si c’est gratuit ou payant, mais où placer l’argent public.” Almost stated, and deliberately placed as the easiest item near the end.
8. b) Critique mais mesurée. The author disagrees, concedes twice, and attacks nobody personally. Option A catches candidates who register the disagreement but miss the concessions. Tests: tone.
Two questions here are worth studying rather than just checking. Question 1 works only if you notice that paragraphs one and two carry the opposing case, not the author’s. Question 6 asks what a paragraph does, and that question type barely exists below B2.
There is a moment in most B2 groups when someone stops asking what a paragraph means and starts asking what it is for. Scores move noticeably after that. It is not a vocabulary threshold; it is a change in how the text is being read, and it is the single reason we teach paragraph labeling before we teach any reading vocabulary at all.
How Do You Read for Paragraph Function?
Give each paragraph a one-word label as you read. Case, evidence, objection, concession, reframe. Six labels cover almost any argumentative text, and they answer functional questions directly.
Applied to the text above:
| Paragraph | Function |
|---|---|
| 1 | Sets up the popular case |
| 2 | Gives the evidence for that case |
| 3 | First objection: capacity |
| 4 | Second objection: financing |
| 5 | Third objection: unintended effect |
| 6 | Concession, limiting the criticism |
| 7 | Reframe, stating the real question |
That table took thirty seconds to build and answers questions 5, 6, and 8 on its own.
Signposts make the labeling easy. Pourtant, or and en revanche open objections. Faut-il, pour autant, certes et il est vrai, que open concessions. D’abord, ensuite, and enfin number the arguments, so you know how many objections are coming.
Practice on any French opinion column. Label the paragraphs before you think about content. Ten minutes a day for a fortnight changes the way you read.
Argument architecture is tested across the whole diploma. The B2 written mock paper asks you to build this structure yourself, and the DELF B2 oral papers have you build the same structure aloud.
Which Vocabulary Carries Nuance at B2?
Les mots qui qualifient un énoncé plutôt que de l’affirmer : nuancer, relativiser, tempérer, écarter, souligner, dénoncer, préconiser, déplorer, se garder de et n’en reste pas moins.
Three groups repay memorizing:
- Verbs of position. Affirmer, soutenir, contester, réfuter, admettre, concéder, reconnaître. Questions about what an author does use exactly these.
- Verbs of degree. Nuancer, tempérer, atténuer, accentuer, exagérer. Question 6 above is answered by knowing what nuancer means.
- Tone adjectives. Mesuré, virulent, ironique, prudent, catégorique, réservé, alarmiste. These fill the options for tone questions. A candidate who cannot separate mesuré from prudent is guessing.
Build the list from the options in practice papers, not from a textbook. The vocabulary in the questions matters as much as the vocabulary in the texts. Candidates rarely study it.
Students in our B2 reading prep course keep a running list of optional vocabulary for this reason. It is the fastest single gain in the paper.
How Should You Use the Hour?
Around eighteen minutes per long text, leaving a few minutes to check. Read the questions first. Then read the text once at a steady pace, labeling paragraphs as you go.
A workable routine:
- Three minutes. Survey the paper. Note which texts are informative and which are argumentative.
- Eighteen minutes per argumentative text. Questions first, one careful read with paragraph labels, answer as you go.
- Twelve minutes for shorter informative documents. These are faster, since they test location rather than reasoning.
- Five minutes. Check anything flagged, prioritizing tone and intent questions.
Never read a 500-word text twice in full. There is no time. A second reading rarely changes an answer a paragraph label has already settled.
Flag rather than linger. If two options both look right, mark the question and move on. Coming back with a fresh eye takes twenty seconds and beats three minutes of staring.
What Costs B2 Candidates Marks?
Answering from paragraphs, the author is refuting, missing a concession, picking the option that quotes the text, and treating an argumentative piece as though it were informative.
Two more patterns show up in mock papers.
- Confusing strength of position. The author criticizes the policy but does not reject it. Options a and b in question 5 differ only by degree, and B2 papers work that seem constant.
- Ignoring the title. “La gratuité des transports, fausse bonne idée ?” is a question, not a statement. The question mark signals a nuanced answer. Titles at B2 usually preview the structure.
Reading and listening are two skills on the B2 test. Follow an argument and identify the position. DELF B2 audio papers cover the audio version, where speakers concede and reverse in the same way.
Why B2 Reading Carries Weight Now
Since 1 January 2026, French naturalization has required B2 in all four skills rather than B1, under law 2024-42 and decree 2025-648. Reading has to reach the level on its own.
B2 is also, with rare exceptions, the level French universities require. For applicants from India, that makes it the working target, not an aspiration.
The reading skill itself is what keeps paying afterward. At B2, you can read French press, professional documents, and academic summaries without translation. That is what studying or working in French demands.
For Canada, the distinction holds as always. Immigration points are based on TEF Canada or TCF Canada, and those results expire after two years.
DELF diplomas never expire. If Canada is the destination, the TEF Canada reading paper shows what that exam asks instead. If B2 is your target level rather than your current level, the time needed to reach it provides a realistic study plan.
Common DELF B2 Reading Questions From Candidates
Sit the Full Paper Before You Sit the Exam
B2 reading is not a vocabulary test. It tests whether you can hold a seven-paragraph argument in your head and say what each part is doing. That is a trained skill, and it needs marked practice.
Our DELF exam preparation runs complete timed reading papers. Then we work through every distractor, so you can see why the plausible option was wrong.
Batches stay at six to ten students at our Noida center. French classes online run the identical papers wherever you are.
Call or WhatsApp 7011164582 to book a free demo and a full-length reading mock.