The 2026 Selective High School Exam: What Parents Need to Know

If you have a Year 4 or Year 5 child in Sydney, you've probably been told ten different things about the Selective exam by ten different people. Your neighbour's cousin who tutored their daughter through it three years ago, the aunty at the temple who swears by one particular centre in Hurstville, the parent at soccer pick-up who whispers that the North Shore centres are "the only real ones." It's confusing because so much of what you're hearing is based on an exam that doesn't exist anymore.

The NSW Selective High School Placement Test changed meaningfully in 2021 and then again — much more significantly — in 2025. Most of the advice circulating in Sydney parent networks is based on the pre-2021 version. I want to write this down clearly because I get asked the same questions every week, and the honest answers save families a lot of wasted money.

What is the Selective test, really?

It's the single assessment that determines entry into all 51 fully and partially selective high schools in NSW — James Ruse, North Sydney Boys and Girls, Sydney Boys and Girls, Baulkham Hills, Hornsby Girls, Normanhurst Boys, Ruse, Penrith, St George Girls, and the rest. Your child sits it once in Year 6 for Year 7 entry. One morning, four sections, 155 minutes on a computer. That's it.

There is no appeal. No interview. No portfolio. The test result plus your child's Year 5 school moderation score produce a single placement score, and that number — ranked against roughly 15,000 other kids doing the same test — determines the outcome. For a place at James Ruse you're looking at top ~200 or so. For somewhere like Baulkham Hills or Hornsby, meaningfully further down. For a "fully selective" offer anywhere, you're still talking top 15% of a very strong applicant pool.

This is a high-stakes, low-margin, single-sitting exam. It rewards preparation. It punishes inconsistency. And the preparation that works in 2026 is not the preparation that worked in 2018.

What changed in 2025 (and why your neighbour's advice is stale)

Three things. All three matter.

1. ACER now writes the questions

From 2025 onwards, the Australian Council for Educational Research took over authoring the Selective test content. If you've heard of AAS, ACER's other well-known assessment, you have a sense of what this means. ACER questions are more aligned with thinking-skills logic than with pure curriculum knowledge. Their Thinking Skills and Mathematical Reasoning questions look and feel like cognitive assessments, not textbook exercises.

If your child has been drilled on Matrix or Pre-Uni worksheets from 2018, a lot of that work is misaligned with what ACER actually asks. The old-style "memorise 200 worked examples and recognise the pattern" approach still helps with core maths fluency, but it's no longer enough. ACER specifically writes questions that punish pattern-matching without understanding.

2. The test is fully digital, delivered on Janison

Janison is the platform. Your child does not write on paper anywhere during the exam. They read passages on screen. They select answers with a mouse. They type the entire Writing section on a keyboard. They flag questions for review using an on-screen button.

This sounds minor. It isn't. I have had students who were top of their Year 5 class on paper, and who scored 20 marks below their potential on a computer-based practice test because they'd never trained on one. Reading on screen is slower and more fatiguing than reading on paper for most children. Typing is slower than handwriting for children who haven't practised. Scrolling and re-scrolling through a long Reading passage consumes time that students used to spend re-scanning with their eyes.

If your child has not sat at least 5 full-length practice tests on a computer under timed conditions before exam day, you are gambling.

3. Writing is 25% of the score, and it's typed

Writing used to be a smaller piece. It's now worth the same as Reading, Maths, and Thinking Skills — a full quarter of the mark. And the kids are typing it on a Janison text box with a 30-minute timer. No spell-check. No grammar-check. No ability to reorder paragraphs by dragging them around the way adults do in Google Docs.

A child who writes beautifully by hand but types 15 words per minute will lose marks they should be earning. Around 30–35 WPM is where your child needs to be to produce a writing response of appropriate length in the time given. This is rarely practised at school. Almost no generic tutoring addresses it.

The four sections, in order

Every Selective test follows the same structure on the same day:

• Reading — 45 minutes, roughly 30 questions. Multiple passages, multiple-choice comprehension questions, cloze passages where words drop out and your child picks from a list, and sentence insertion tasks. More about inference and synthesis than "find the answer in paragraph 3."
• Mathematical Reasoning — 40 minutes, 35 questions. Not the Year 5 syllabus. Word problems, multi-step applied maths, geometry, probability, data interpretation, pattern recognition. Heavily reasoning-based.
• Thinking Skills — 40 minutes, 40 questions. ACER-style. Deductive logic, syllogisms, spatial reasoning, argument evaluation, pattern matrices. This is the section that shocks parents who've never seen it before.
• Writing — 30 minutes, one prompt, typed on screen. Could be narrative, persuasive, or discursive. Students don't know which until the day.

All four weighted equally at 25% each. One weak section can tank the whole result.

What 2026 preparation actually looks like

Here's where I'm going to be honest about what works and what doesn't, because I've watched a lot of families throw money at the wrong things.

What doesn't work

• Endless paper worksheets. Necessary for fluency, but insufficient. A child can do 500 paper questions and still fail on computer-based tests if they've never trained on a screen.
• "Past papers" from before 2025. Structurally different exam, differently weighted, written by different authors. Still useful for skill-building, but do not treat old papers as a roadmap.
• Huge classes of 30+ students. You're paying for volume, not attention. Your child's specific weaknesses won't get identified, let alone addressed.
• "Trick" techniques that guarantee a result. There are no tricks. There is systematic preparation and there is hope, and only one of those two things works.

What actually works

• A proper diagnostic. Before you spend a cent, you need to know where your child actually is across the four sections. The gaps are not uniform. A child might be 18 months ahead in maths and 6 months behind in Thinking Skills, and the response to each is different.
• Skill-by-skill teaching, not topic-by-topic. The exam doesn't test "fractions." It tests specific applications of fractional reasoning to word problems under time pressure. Those are different things. Good prep teaches the specific skill.
• Computer-based practice, repeatedly. Not once or twice — five to ten full-length digital mock exams before the real one, so that the format and fatigue stop being a factor.
• Typing practice. Fifteen minutes a day for three months gets almost any primary-aged child to 30+ WPM. It's the single highest-ROI thing you can do for the Writing section.
• Error analysis. After every mock test, the question isn't "what did you get wrong" — it's "what skill was the error testing, and how do we train that skill specifically?" Most tutoring skips this step entirely.

The timeline that works

If your child is sitting the test in Year 6 next year, you have roughly 12 months. Here's the rough shape of how serious preparation looks:

• 12–9 months out: Foundation. Identify skill gaps across all four sections. Shore up mental arithmetic, reading stamina, writing mechanics, typing speed.
• 9–6 months out: Skill acquisition. Systematic coverage of each question type. Introduction to ACER-style Thinking Skills. First timed practice questions.
• 6–3 months out: Application. Full sections under timed conditions. First full-length computer-based mock. Error analysis becomes routine.
• 3–0 months out: Polish and stamina. Five to eight full-length mock tests on the actual Janison-style platform. Strategy refinement. Exam-day routines rehearsed.

If you're starting later than 12 months out, it's not hopeless — I've seen strong late starters succeed — but the plan has to be more intensive and the diagnostic has to be brutal about what to prioritise.

What parents should actually ask any tutor before enrolling

1. How do you deliver computer-based practice? (If the answer is "we use past papers", walk away.)

2. How do you identify which specific skills my child is weak in? (If the answer is generic, the teaching will be too.)

3. What is the class size and student-to-teacher ratio? (Anything over 15 is a worksheet factory, not teaching.)

4. Who teaches? Are they qualified? (A lot of cram-school tutoring is delivered by HSC students or recent grads reading from a manual. Ask.)

5. Have you coached students on the post-2025 ACER/Janison format specifically? (Many centres are still teaching the old exam.)

The hardest thing to hear

Not every child who prepares will get into a selective school. The maths don't work that way — only the top 15% of roughly 15,000 applicants place, and that pool self-selects to be very strong. If you're only preparing because you want a selective outcome at any cost, both you and your child will have a rough 12 months.

The parents I see get the best results treat selective preparation as one of the reasons, not the only reason. They want their child to develop rigorous thinking habits, solid maths, confident typed writing, and the stamina to sit and focus for 155 minutes. Those are gifts that last a lifetime, regardless of whether the selective placement comes through. Keep that frame, and the experience is a good one for your child either way.