Reading / English
PD3 Reading — Exam Archive 2004–2024 (English)
42 exam sittings · Delprøve 1 + 2A/2B · English edition
2 free sample sittings below in full — unlock to read all 42 exam sittings.
Strategy GuideFREE
Read this first. These rules and methods apply to every exam sitting from 2004–2024. Understanding the scoring conventions and recurring traps before you work through the archive will prevent avoidable point loss from the start.
§1 How the Scoring Works
All answers must come directly from the text. The official answer key uses these conventions:
| Symbol / form | Meaning | Practical rule |
|---|---|---|
| (parentheses) | Optional — include or omit, both earn full marks | (De to) energikriser → "energikriser" alone is fine |
| A / B (slash) | "Or" — either phrasing earns full marks | Write one; writing both is fine but not required |
| No slash, no parentheses | All listed components required | If key lists two items joined by "and," both must appear — see §2 |
| 2 or 0 | L2A open questions and L2A multiple choice | All-or-nothing — no partial credit for a half-right answer |
| 1 or 0 | L1 questions and L2B cloze gaps | Same rule — one point per question, all or nothing |
Copy, don't paraphrase. Use the text's own words. A correct paraphrase can still be marked wrong if a specific term was required.
§2 The "Must Write Both" Trap
Some open-answer questions require two (sometimes three) separate pieces of information joined by "and," not "or." Writing only one correct piece scores zero — PD3 does not award partial credit on these.
How to catch it before you lose the point:
- If the question uses hvilke to (which two), to grunde (two reasons), or to betingelser (two conditions) — assume two components are required until confirmed otherwise.
- After writing your answer, re-read the relevant paragraph and ask: "Did the text give me more than I wrote?" If yes, you probably missed the second piece.
- The dangerous case: the official key lists two items with no slash, or uses explicit wording that both are mandatory.
- Distinguish this from a slash (/) in the key, which always means "either is fine."
Classic example — Summer 2010, Reading 1 Q10: The question asks which authorities to contact. The key requires both Arbejdstilsynet and Arbejdsskadestyrelsen — writing only one scores 0.
Three signals that a question requires two pieces:
- The question uses hvilke to, to grunde, or nævn begge: "Nævn begge" (name both) is the clearest signal on the exam paper — treat it as a direct instruction to write two items, never just one.
- The question asks for types of activities, materials, or names: If the text lists two nouns joined by "og" (and) with no slash between them, both are required. Writing only one, even if correct, scores zero.
- The question asks "which two words describe X": Review and evaluation questions often require two adjectives simultaneously — distinguish from the slash (/) case where only one is needed.
Known historical examples (for review): Summer 2010 Q10 (two agency names) / Cultural event double activity (city walks og archaeological excavations) / Craft material double item (glass og ceramics) / Reviewer double adjective — every case shares the same marker: the key lists two items joined by "og" with no slash.
Build the habit: After writing any open answer, ask yourself: "Did the text actually give me two pieces of information here?" This check takes five seconds and catches the most expensive mistakes.
§3 Connector Reference Table
Learn the logical relation each connector signals, not just its translation. Exam gaps test whether you understand what the sentence is doing, not whether you know the dictionary definition.
| Danish word(s) | Logical relation | What to look for in context |
|---|---|---|
| selvom / skønt / trods | Concession — "even though / despite" | "Even if X, still Y." Gap often sits in the first half — reading only the second half misses the concessive signal entirely. |
| alligevel | Contrast — "nevertheless / still" | The text just stated a fact or condition; this word signals the outcome is contrary to what that fact would predict. |
| imidlertid / dog / men | Contrast — "however / but" | Standard pivot to counterpoint. Slightly more formal than men. |
| hvorimod / derimod | Direct contrast — "whereas" | Find the two items being directly opposed. Usually parallel subjects on either side of the connector — more formal than men. |
| desuden / endvidere / ligeledes | Addition — "furthermore / moreover" | A new point is being added to what was just said, not contrasting it. |
| nemlig | Confirms / explains — "namely / in fact" | The previous sentence made a claim; nemlig introduces the specific fact that proves or explains it. Look backward. |
| faktisk | Surprising fact — "actually" | Signals the upcoming clause contradicts what a reader would expect — counterintuitive result. |
| tilsammen | Joint effect — "together / combined" | Two previously-separate factors are now said to work together. Find what's being combined — they were named earlier in the text. |
| dermed / derfor / således / derved | Result — "thereby / therefore / thus" | The preceding clause is the cause; this connector introduces the direct consequence. |
| hvis / medmindre | Condition — "if / unless" | Medmindre is the negative form ("unless") — easy to confuse with hvis in multiple-choice options. |
Double-negation trap: A gap near ikke is easy to misread. "ikke så dårligt" means "not so bad" — a mild positive. Slow down on any gap where ikke appears nearby.
§4 L1 Strategy: Reading 1 (15 questions, 1 pt each)
L1 texts are information documents — guides, catalogs, brochures, official leaflets. Questions follow a predictable pattern:
- Hvad / Hvem / Hvor → direct fact — find the matching noun or name
- Hvornår / Hvor lang tid → time or duration — look for numbers, dates, period markers
- Hvorfor → reason — find fordi / da / skyldes / forklarer and the clause that follows
- Hvordan → method or mechanism — find process description or comparison
- Ja / Nej → check the exact wording — "or" (eller) means neither option is guaranteed, which often flips the answer
Anchor-word technique: Find a specific word from the question that also appears in the text — that sentence is almost always where the answer lives.
Worked examples — L1 four question types in action
| Type | Example question (simplified) | Steps | Answer |
|---|---|---|---|
| Hvad/Hvem What / Who | Hvad er Helligdomsklipperne opkaldt efter? (What is Helligdomsklipperne named after?) | ① Circle anchor: Helligdomsklipperne ② Find that word in the text ③ Extract the noun explaining the origin of the name | (En hellig) kilde Parenthesised word optional |
| Hvornår When | Hvornår fandt man ud af, hvornår Hammershus er bygget? (When was the construction date of Hammershus established?) | ① Dual anchor: Hammershus + fandt ud af ② Find the year number co-occurring with both anchors ③ Watch the trap: the question asks when the date was discovered, not when the castle was built | I 2015 |
| Ja/Nej Yes / No | Er der en udstilling om helligdomskilden på kunstmuseet? (Is there an exhibition about the holy spring at the museum?) | ① Anchor: udstilling (exhibition) + Kunstmuseum ② Text says the spring flows through the building as an architectural feature — not an exhibition ③ Trap: both words appear near each other, tempting a "Ja" | Nej |
| Two-part answer Hvilke / Hvem | Hvilke to ord bruger anmelderen om romanen? (Which two words does the reviewer use about the novel?) | ① Anchor: anmelderen (reviewer) ② Find the reviewer's evaluation sentence in the text ③ Confirm there are two adjectives joined by "og" (no slash) ④ Write both — missing one scores zero | Troværdig og genkendelig Both words required |
Format reminder: L1 answers should be concise — typically 3–8 words, no full sentence required. Extract noun phrases or adjectives directly. Words in parentheses in the official key (e.g. (En hellig)) are optional: including them does not lose marks, omitting them does not either.
§5 L2A Strategy: Open Questions (2004–2013 format)
When L2A requires you to write a Danish answer (not select A/B/C/D), use the same anchor-word technique as L1, plus:
- Multi-perspective articles: When 2–4 speakers are presented, identify whose opinion each question asks about before scanning for the answer. Mixing up speakers is the most common error on these questions.
- Inference questions: Look for phrases that express attitude — glæder sig / frygter / mener / understreger. Inference is not guessing; the attitude must be stated in the text.
- Cause-effect questions: The answer clause usually starts with fordi / da / som følge af or is followed by derfor / dermed.
§6 L2A Strategy: Multiple-Choice Questions (2014–2024 format)
From 2014 onward, every L2A question is A/B/C/D multiple choice, 2 points each, all-or-nothing (no partial credit).
- Read all four options before touching the text. Knowing what the distractors claim tells you exactly which details to hunt for.
- Skim the article once for structure — who is speaking, how many paragraphs, where the topic shifts — before answering any question.
- Locate the sentence(s) that directly support or contradict each option. Every official answer traces to specific text; there are no answers based on common sense outside the text.
- Eliminate first. Rule out any option the text directly contradicts — this is almost always faster than confirming the correct answer from scratch.
- Between two plausible options, pick the one with the more precise and specific textual match, not the one that merely sounds reasonable in your own language.
All-or-nothing reminder: "Feels right" without textual support is worth exactly the same as a wrong answer. Don't skip step 3.
Worked example — dissecting a multiple-choice question option by option
Text background (simplified): Two nurses, Mette and Henriette, both became ward nursing managers in their thirties. Mette expected managing former colleagues to be hardest; what actually proved hardest was having no peer at the same level to talk to. Henriette's manager gave her no clear framework of responsibilities, and she ended up carrying everything alone, eventually going on extended sick leave. Both women admitted they initially found it difficult to delegate tasks.
Question: Hvad var den sværeste udfordring for Mette, da hun startede som afdelingssygeplejerske?
(What was the hardest challenge for Mette when she started as ward nursing manager?)
| Option | Right / Wrong | Distractor type and analysis |
|---|---|---|
| A. Det var svært at lede sine gamle kolleger Managing her former colleagues was difficult | ❌ | "Expectation vs. reality" trap: Mette expected this to be hardest — but the text uses a contrast word (faktisk / men / imidlertid) to signal that something else actually was. Stopping before the contrast word and selecting A is the most common error on this question type. Always read the full sentence. |
| B. Hun manglede en kollega på samme niveau at dele erfaringer med She lacked a peer at the same level to share experiences with | ✅ | The text directly states this as Mette's main difficulty. Anchor words: Mette + samme niveau (same level) — find where both appear together in the same sentence. |
| C. Hendes leder gav hende ikke en klar jobbeskrivelse Her manager did not give her a clear job description | ❌ | Speaker-confusion trap: This is what happened to Henriette, not Mette. The single most common distractor in multi-speaker articles: the writer takes one person's experience and attributes it to another. Always circle which person the question is about before scanning the text. |
| D. Hun havde svært ved at uddelegere arbejdsopgaver She found it hard to delegate tasks | ❌ | Scope-inflation trap: The text does say both women struggled with delegation — but the question asks for Mette's hardest challenge (sværeste udfordring), and delegation is described as a shared secondary weakness, not her main difficulty at the start. Partly true ≠ correct answer. |
Elimination order: First cut C (wrong person) → cut A (expectation reversed by contrast word) → cut D (true detail but wrong scope) → confirm B has direct textual support. Target: resolve a 2-point MC question in 3–4 minutes.
§7 L2B Strategy: Cloze / Gap-Fill (8 gaps, 1 pt each)
Every Reading 2 paper from 2007 onward includes a second text with 8 numbered gaps — multiple choice in most years, occasionally free-fill in 2007–2010. Worth as much as half of L1. Don't save it for the last five minutes.
The Four-Step Method (Grammar first, meaning second, logic eliminates the rest):
- Identify the word class the gap needs. Is the missing word a noun, verb, adverb, or connector? This alone eliminates grammatically impossible options before you consider meaning.
- Read both sides of the gap. What does the sentence immediately before say? What does the sentence immediately after say? The correct answer must make both halves connect logically. Reading only one side is the most common cause of picking a plausible-sounding wrong answer.
- Identify the logical relationship, then match it to the connector type using the table in §3 above.
- Eliminate grammatically impossible options first, then choose the best semantic fit among what remains. Don't judge grammar and meaning simultaneously — it's slower and more error-prone.
Concessive structure trap: A sentence shaped "even if X, still Y" (selvom-clauses) often puts the gap inside the first half. If you only read the second half for clues, you'll miss the concessive signal entirely.
§8 Exam Timing
Recommendations below are based on the common exam format — scale proportionally if your actual session runs longer or shorter.
| Section | Typical volume × points | Suggested time | Pacing note |
|---|---|---|---|
| Reading 1 (L1) | 15 questions × 1 pt | 25–30 min | ~1.5–2 min per question. Hvad/Hvem questions are fastest; Ja/Nej questions can drag — cap at 2 minutes, mark and move on. |
| Reading 2A (L2A) | 3–5 questions × 2 pts | 15–20 min | Read all four options before touching the text. If stuck for more than 4 minutes, skip and return — a 2-point question doesn't justify unlimited time. |
| Reading 2B (L2B) text-insertion | 5–8 gaps × 1–2 pts | 10–15 min | Identify word class first, then read both sides of the gap. Never use only one side as your evidence. |
| Cloze / connector gaps | 8 gaps × 1 pt | 8–12 min | Word class → §3 table → eliminate → confirm. About 1 minute per gap. |
| Final check | — | 5 min | Two things only: ① did you miss a second information point anywhere (§2 trap)? ② does every answer come from the text rather than background knowledge? |
Priority rule: Reading 1 has the most stable points and the fastest pace — complete it before moving on. L2A is all-or-nothing, so skipping a stuck question beats spending 8 minutes on it. The cloze section is only 8 gaps at about 1 minute each — never abandon it to save time.
The most common time leak: Spending too long on Reading 1 when it's going well, then rushing L2A. Reading 1's goal is fast locate-and-extract, not thorough comprehension — move quickly and keep the bigger-point sections protected.
The archive below contains complete question-by-question analysis for every exam sitting 2004–2024. Apply these techniques as you work through each session.
Maj/Juni 2024FREE
📅 Summer 2024 (Maj/Juni 2024)
Format: Reading 1 (15 questions, 1 pt each) + 2A (3 MC, 2 pts each) + 2B text-insertion (5 gaps, 2 pts each) + Delprøve 3 cloze (8 gaps, 1 pt each) = 39 pts total.
Reading 2B: En fire-dages arbejdsuge? (A Four-Day Work Week?) — Text Insertion
Format: Five gaps (1–5) in the text. Seven fragments (A–G) — five fit, two are distractors.
Scoring: 2 pts each, all-or-nothing. ✅
Fragments
- A. I Danmark er der nemlig hver dag 35.000 sygemeldinger som følge af stress. Dertil kommer selvfølgelig andre sygemeldinger og andet fravær på grund af almindelig nedslidning.
- B. Det forventes nemlig som nævnt, at nye medarbejdere ikke arbejder hjemme, men altid møder ind på arbejdspladsen.
- C. Der er dog udfordringer ved individuelt tilpassede arbejdsforhold, da medarbejdere har forskellige behov. Nogle trives for eksempel bedst med at arbejde på kontoret, andre hjemme. Nogle foretrækker at arbejde tidligt, andre sent.
- D. Måske fordi det umiddelbart kan virke ulogisk at nedsætte arbejdstiden i en tid, hvor der er mangel på arbejdskraft. Men det er det ikke nødvendigvis.
- E. Den daglige drift er derfor ledernes ansvar, men det er stadig medarbejderne, der udfører opgaverne. Og tilfredse medarbejdere genererer mere værdi i virksomheden.
- F. I nogle virksomheder er man lykkedes med at lære nye og mere effektive arbejdsrutiner. Man udfører det samme arbejde på kortere tid, og lønnen er forblevet den samme.
- G. Når flere virksomheder indfører en fire-dages arbejdsuge, skal kunder og borgere også vænne sig til, at man ikke kan komme i kontakt med en medarbejder fem dage om ugen.
Official Answer Key
| Gap | 1 | 2 | 3 | 4 | 5 |
|---|---|---|---|---|---|
| Answer | C | A | F | D | G |
Gap-by-gap coherence logic
- Gap 1 → C: Preceding text introduces the idea that companies want to make working conditions more flexible and tailored to each employee. C immediately names the complication: "Der er dog udfordringer" (There are however challenges) — different employees need different things (office vs. home, morning vs. evening). Cohesion signal: dog (however) + challenge — C is the counterweight to the positive proposal just made.
- Gap 2 → A: Preceding text says the four-day week has attracted much attention, and "there may be good reason for this." A begins with nemlig — "namely" — which introduces the specific evidence confirming that claim: 35,000 Danish sick days per day from stress. Cohesion signal: nemlig as causal confirmation — "good reason": here is the concrete data that proves it.
- Gap 3 → F: Preceding text states that 37 hours of work can be completed in 4 days instead of 5. F explains the mechanism: "some companies have succeeded in learning new and more efficient work routines — same work in shorter time, same salary." Cohesion signal: mechanism elaboration — the preceding claim ("it can be done in 4 days") is followed by F explaining how companies actually achieved it.
- Gap 4 → D: Preceding text acknowledges that there is still resistance from some employers. D explains why: "Måske fordi det umiddelbart kan virke ulogisk at nedsætte arbejdstiden i en tid, hvor der er mangel på arbejdskraft" (Maybe because it can seem illogical to reduce hours during a labor shortage). Cohesion signal: causal explanation for resistance — "maybe because" directly answers the implied question of why employers resist. The second sentence ("Men det er det ikke nødvendigvis" = but it's not necessarily illogical) adds a rebuttal that keeps the article's balanced tone.
- Gap 5 → G: Preceding text says more companies are offering the four-day week or signaling they will. G draws the societal consequence: "when more companies adopt a four-day week, customers and citizens will also have to get used to not being able to reach a staff member five days a week." Cohesion signal: societal ripple effect — the article widens from companies to the broader impact on service users.
Why B and E are distractors (cannot fit any gap)
- B is eliminated: B says new employees are expected to come to the office rather than work from home. This is about remote-work expectations for new hires — a specific sub-topic that doesn't fit any gap. Gap 1 needs a challenge about individual needs (C handles this more broadly), and B's narrow claim about new employees' office requirement has no contextual home in the five positions.
- E is eliminated: E says daily operations are management's responsibility, but employees execute the tasks — and satisfied employees generate more value. This is about management-worker division of responsibility — a thematically adjacent but positionally homeless observation. No gap in the article calls for this management-theory insertion at the right point in the argument flow.
Delprøve 3: Nethandel og storcentre truer livet i bymidterne (Online Shopping and Shopping Centres Threaten Town Centres) — Cloze
Article structure: Online shopping and out-of-town malls → empty town-centre shops (problem) → high rents worsen it → government ten-town trial to reimagine town centres → local shops can compete by matching online service → their unique advantage: historic buildings and atmosphere.
Scoring: 1 pt per gap, 8 gaps. ✅
Answer Choices & Official Key
| Gap | A | B | C | D | Answer |
|---|---|---|---|---|---|
| 0 (example) | investere | åbne | lukke ✓ | udvide | C |
| 1 | omvendt | alligevel | derfor | dog | C |
| 2 | ikke alene | derudover | tværtimod | næsten aldrig | A |
| 3 | omkostninger | forsikringer | huslejer | udbetalinger | C |
| 4 | et mål | et argument | en undtagelse | en tendens | D |
| 5 | varernes kvalitet | velstanden | udvalget af varer | livskvaliteten | D |
| 6 | imidlertid | desuden | derimod | nemlig | B |
| 7 | service | rente | vejledning | reklame | A |
| 8 | et ansvar | en fordel | et mål | en udfordring | B |
Gap-by-gap reasoning
Example gap 0 — lukke (C): Shops in town centres _____ (close). The article opens by naming the problem: empty storefronts. All other options (investere = invest, åbne = open, udvide = expand) imply growth — the article's premise is the opposite.
Gap 1 — derfor (C): Customers go online or to malls → _____ more storefronts are empty. Derfor (therefore) is the result connector — cause-and-effect chain. B (alligevel = nevertheless) would imply the storefronts empty despite something, adding an unexpected-outcome element the sentence doesn't have. D (dog = however) signals contrast — but cause and effect here run in the same direction. A (omvendt = conversely) signals reversal — also wrong direction.
Gap 2 — ikke alene (A): The problem is _____ online competition; high rents also contribute. Ikke alene (not only) introduces a second factor to a multi-causal problem: "not only X but also Y." B (derudover = additionally) could also work grammatically to add a new factor. The distinction: ikke alene specifically signals "you might think this is the only cause — but here's another one," which is the article's rhetorical move at this point. C (tværtimod = on the contrary) signals direct opposition — high rents are not the opposite of online competition. D (næsten aldrig = almost never) is an adverb of frequency, wrong word class.
Gap 3 — huslejer (C): Property owners in town centres charge high _____ for their spaces. Huslejer (rents) is the specific economic charge that worsens the situation. A (omkostninger = costs) is too vague — "costs" could be anything; the text is specifically about what landlords charge. B (forsikringer = insurance premiums) is a real business cost but not the article's topic. D (udbetalinger = payouts/payments) lacks a clear referent for "payments from whom to whom" in this context.
Gap 4 — en tendens (D): The decline of town-centre trade is _____ that worries many politicians. En tendens (a trend) — a developing observable pattern over time that justifies policy concern. A (et mål = a goal/target) would mean declining trade is what politicians aim for — logically backwards. C (en undtagelse = an exception) would frame declining trade as unusual/isolated — the article says it is widespread and ongoing. B (et argument = an argument) could work grammatically but "a trend" is more precise for a persisting observable phenomenon than "an argument."
Gap 5 — livskvaliteten (D): When all local shops close, it can harm the _____ of residents. Livskvaliteten (quality of life) — empty town centres mean fewer accessible services, social spaces, and local convenience for everyday life. A (varernes kvalitet = quality of goods) — closing shops means no goods at all, not lower quality goods; the concern is absence, not degradation. C (udvalget af varer = range/selection of goods) — again about goods specifically, not about living conditions overall. B (velstanden = wealth/prosperity) — the article's concern is about livability and community wellbeing, not economic wealth in general.
Gap 6 — desuden (B): The ten selected towns have _____ incorporated green areas and event spaces into the reimagined town centre plans (in addition to welfare services just mentioned). Desuden (furthermore/also) adds a new element in the same positive direction. A (imidlertid = however) signals contrast — but adding green areas is in the same direction as welfare services, not a reversal. C (derimod = on the other hand/whereas) sets up a binary opposition that isn't present here. D (nemlig = namely/in fact) would say green areas confirm or explain the welfare services — but green areas are an additional element, not an explanation of welfare services.
Gap 7 — service (A): Local shops can offer the same _____ as online shops — delivery options, price matching. Service matches the operational capabilities that online shopping is known for and that local shops can replicate. B (rente = interest rate) is a financial term — completely unrelated to retail. C (vejledning = guidance/advice) — local shops might offer expert advice (and often do), but the specific comparison being made here is about logistical services (delivery, price matching) that online shops are famous for. D (reklame = advertising/marketing) — matching online advertising is not what the article's advice is about.
Gap 8 — en fordel (B): Local shops have _____ that neither online shopping nor modern malls can replicate: historic buildings and atmosphere. En fordel (an advantage) — this is the article's final positive argument for local retail's unique value. A (et ansvar = a responsibility) would frame historic buildings as a burden or obligation for local shops — opposite of the intended positive argument. D (en udfordring = a challenge) similarly frames historic buildings as a problem. C (et mål = a goal/target) would mean shops aim to have historic buildings, making no sense as a description of an existing competitive asset.
- Maj/Juni 2023
- Nov/Dec 2023
- Maj/Juni 2022
- Nov/Dec 2022
- Maj/Juni 2021
- Nov/Dec 2021
- Maj/Juni 2020
- Nov/Dec 2020
- Maj/Juni 2019
- Nov/Dec 2019
- Maj/Juni 2018
- Nov/Dec 2018
- Maj/Juni 2017
- Nov/Dec 2017
- Maj/Juni 2016
- Nov/Dec 2016
- Maj/Juni 2015
- Nov/Dec 2015
- Maj/Juni 2014
- Nov/Dec 2014
- Maj/Juni 2013
- Nov/Dec 2013
- Maj/Juni 2012
- Nov/Dec 2012
- Maj/Juni 2011
- Nov/Dec 2011
- Maj/Juni 2010⚠️ Reading texts missing — answer key verified, but exam articles not in collection
- Nov/Dec 2010
- Maj/Juni 2009⚠️ Incomplete — question maps only, most answers not recovered
- Nov/Dec 2009⚠️ No answer key — question headings only, answers not recovered
- Maj/Juni 2008⚠️ No answer key; L1 text partial — L1 frameworks only, L2 complete
- Nov/Dec 2008⚠️ Incomplete — most Reading 1 answers and Reading 2B cloze missing
- Maj/Juni 2007⚠️ No answer key; L1 text missing — L1 unverified, L2 complete
- Nov/Dec 2007
- Maj/Juni 2006
- Nov/Dec 2006
- Maj/Juni 2005⚠️ Reading texts + answer key both missing — question frameworks only
- Nov/Dec 2005
- Maj/Juni 2004
- Nov/Dec 2004⚠️ Reading texts + answer key both missing — question frameworks only
Unlock all exam sittings (2004–2024) — one-time payment, lifetime access.
Unlock — 119.9 krPay with the email you want to use — we'll automatically email you a sign-in link the moment payment goes through.