The Complete HKDSE ICT Preparation Guide (2026)
This guide is for any HKDSE ICT student who wants a clear roadmap from today to exam day. It reflects the revised syllabus in force since the 2025 HKDSE (unchanged through 2027) and real patterns from past papers. One warning before you start: pre-2025 papers follow the old exam format. They are still excellent for topic drilling — but learn the exam format from 2025+ papers. You’ll leave with a concrete 12-week plan and know exactly how to allocate your limited time.
The exam at a glance
Paper 1 (Compulsory Part, 55%, 2 hours): Section A: 40 multiple-choice questions (22% of the subject mark). Section B: short and structured questions (33%). All questions compulsory. No penalty for wrong answers — answer every question.
Paper 2 (Elective Part, 25%, 1 hour 30 minutes): Answer TWO of the three electives, all questions on both:
- 2A Databases — SQL, ER diagrams, normalisation
- 2B Web Application Development — HTML, CSS, JavaScript, PHP, web programming, networking basics
- 2C Algorithm and Programming — algorithms, Python programming, data structures
Under the revised curriculum (first examined at the 2025 DSE), the old four-elective menu — including 2D Multimedia — no longer exists. Your school teaches two of the three electives, and that choice effectively determines your Paper 2.
School-based Assessment (20%): Two guided tasks completed in class — Task 1 Design & Implementation (25 marks) and Task 2 Testing & Evaluation (15 marks). Private candidates are exempt: their grade comes from the written papers alone.
Paper 1 topic map and rough weighting
The compulsory part has five major areas. Here’s what appears in past papers:
| Area | Core Topics | Typical Past Paper Focus |
|---|---|---|
| Information Processing | Data vs information; ASCII/Unicode (7-bit vs 16-bit); Validation vs verification (type check vs double entry); Binary representation of characters | Data collection methods (DSE 2015 Paper 1 Q5 asked about RFID tag storage size); Parity-bit error detection (DSE 2015 Paper 1 Q1) |
| Computer System Fundamentals | CPU architecture (CU, ALU, registers, fetch-decode-execute); Memory types (RAM vs ROM vs cache); Storage (HDD vs SSD); System vs application software | CPU component that fetches instructions (DSE 2015 Paper 1 Q13); Memory hierarchy and access speed comparisons appear frequently |
| Internet & its Applications | LAN vs WAN; Network topologies (star/bus/ring/mesh); TCP/IP; IPv4 vs IPv6; DNS; HTTP vs HTTPS; HTML basics | HTML vs text format advantages (DSE 2015 Paper 1 Q19); Network protocol questions recur heavily |
| Social Impacts | Ethics; Intellectual property (copyright, patents, software licensing); Privacy & security (phishing, malware, firewalls, encryption); Digital divide; Health & ergonomics (RSI, ergonomics) | Social impact questions appear in every paper; phishing email recognition (DSE 2015 Paper 1 Q38); Digital divide definition (DSE 2016 Paper 1 Q36-Q37) |
| Computational Thinking & Programming | Flowchart symbols and structure; Pseudocode basics; Trace tables; Algorithm representation; Basic program code reading (JavaScript-style) | Flowchart interpretation questions appear regularly; Trace table exercises in multiple past papers; Algorithm efficiency comparisons |
Weighting reality: The official syllabus doesn’t publish exact percentages. All five areas recur heavily in past papers — focus on mastering each rather than memorizing invented percentages.
Your electives, compared
Paper 2 is worth 25% — a full quarter of your grade sits on the electives. Here’s a practical comparison:
| Elective | What You Learn | Who It’s For | Availability |
|---|---|---|---|
| 2A Databases | SQL (SELECT, JOIN, GROUP BY, HAVING); ER diagrams; Normalization (1NF, 2NF, 3NF); Data integrity | Students who enjoy logic and structured problem-solving | Widely taught — the most common elective |
| 2B Web Application Development | HTML/CSS; JavaScript; PHP; Form handling; Web programming; Networking basics | Students interested in building websites and web apps | Depends on the school |
| 2C Algorithm and Programming | Sorting (bubble, selection, insertion, merge); Searching (linear, binary); Algorithm efficiency; Data structures (stacks, queues); Python programming | Students with strong math/logic; programming mindset | Less common — needs a programming background |
My advice: Find out which two electives your school teaches — that answers most of the question for you. If you get a genuine choice: 2A Databases is usually the safest, most structured, exam-friendly pick; 2C rewards students who already enjoy programming but demands more practice; 2B suits hands-on web builders. Whichever two you sit, weight your revision time evenly across both — Paper 2 marks them equally, and students routinely over-focus on one elective and leak marks in the other.
A 12-week study plan
Here’s a concrete timeline from diagnostic to exam-ready. Adjust based on your starting point, but preserve the sequence.
| Phase | Weeks | Focus | Practice Drills |
|---|---|---|---|
| Diagnostics + HA | 1-2 | Take a full past paper under timed conditions. Identify weak areas. Complete all Information Processing (HA) topics from notes. | Past paper MC (40Q timed) + structured Q practice; Error log (track question type, not just topic) |
| HB — Computer Systems | 3-4 | CPU architecture, memory/storage, binary/ASCII, system software. Focus on fetch-decode-execute cycle diagrams. | MC drills + structured Q on CPU components; Memory comparison tables (RAM/ROM/Cache) |
| HC — Internet & Apps | 5-6 | LAN/WAN, topologies, TCP/IP, DNS, HTTP vs HTTPS, HTML tags. Draw network diagrams from memory. | Protocol matching drills; HTML tag recall; Network topology comparison |
| HD + HE — Social Impacts | 7-8 | Ethics, IP rights, security threats (phishing, malware, firewalls, encryption), digital divide, ergonomics. | Case study analysis: “What security principle is violated?” Real-world scenario Qs |
| Elective Deep Dive | 9-10 | Cover both electives your school teaches. For 2A: SQL clause order, ER diagram rules, normalization violations. For 2B: web programming patterns. For 2C: algorithm tracing and Python practice. | Past paper elective Qs (timed, 2025+); Build your own “common mistake” checklist |
| Timed Past Papers + Error Log Review | 11-12 | Full papers under exam conditions. Review your error log — prioritize recurring mistake patterns. | 2-3 full past papers/week; Each mistake: rewrite the correct solution in your notebook |
Phase transition rule: Don’t move to the next phase until you can score 70%+ on mixed-topic practice questions from the current phase. Quality over speed — a solid foundation prevents later panic.
How marks are actually lost
After grading hundreds of papers, the same mistakes repeat. Here’s what kills scores:
1. Misreading command words: “State” means 1-2 marks, simple facts. “Explain” needs 2-3 sentences showing cause-and-effect. “Compare” requires similarities AND differences. You lose 1-2 marks per question when you under-answer.
2. Missing units or justification: Numeric answers without units = half marks. SQL answers without explaining the clause choice = partial marks. Always ask: “Would this convince a marker?”
3. SQL clause-order mistakes: The order matters — SELECT → FROM → WHERE → GROUP BY → HAVING → ORDER BY. One wrong clause position = 0 marks for the entire question.
4. MC time pressure: Budget roughly 50 minutes for the 40 MC questions (about 1.25 minutes each), leaving about 70 minutes for Section B. Students who spend 4+ minutes on one MC question starve Section B of time. Mark it and move on — return later if time permits.
5. Panic on unfamiliar scenarios: Social impact questions often describe real-world situations (e.g., “A hospital implements patient records online…”). Don’t panic — map the scenario to syllabus concepts you know (privacy, security, access control).
Internal guides: For deeper dives on command words and SQL survival, see our command words guide and SQL survival guide.
Where this fits in your prep
This guide gives you the roadmap. Our HKDSE ICT notes provide the detailed content for each topic. Combine them: use the notes for deep learning, use this guide for strategic planning and timing.
One final truth: you can’t self-study your way to a Level 5 without feedback. You need someone to review your structured answers, correct your SQL logic, and calibrate your pacing. That’s what tutors do — we shorten the feedback loop so you don’t practice mistakes for weeks.
Start with a diagnostic paper this week. Build your error log. Follow the 12-week sequence. Check in every 2 weeks with timed practice and targeted review. You’ll see progress by Week 4 if you’re consistent.
Ready to start? Book a trial lesson — we’ll review your diagnostic paper together and build your personalized plan.
Keep going
Every spoke of this cluster: