How to Pass the SHL Numerical Reasoning Test: 12 Expert Tips

The most common mistake candidates make before sitting the SHL Numerical Reasoning test is over-preparing for the wrong thing. They revise long division, brush up on algebra, and practice mental arithmetic — and then arrive at the test to find that it barely resembles any of that.

The SHL Numerical Reasoning test measures data interpretation, not mathematical knowledge. Every question gives you a table, bar chart, line graph, or pie chart, and asks you to extract the right numbers, apply a simple calculation, and select the correct answer from five options. The maths involved is GCSE-level at most: percentages, ratios, currency conversions, and growth rates.

What makes the test genuinely difficult is not the calculation itself — it is the combination of unfamiliar data formats, distracting irrelevant numbers, and strict time pressure. A question that takes 90 seconds on a practice sheet with no time limit becomes a different challenge when you have 65 seconds and 20 more questions waiting behind it.

Common data types you will encounter include: multi-column data tables (often with values in thousands or millions), stacked bar charts, line graphs tracking multiple variables over time, and pie charts requiring proportion calculations. The key skill is knowing which cell of the table or which bar of the chart contains the number you actually need — while ignoring everything else.

The SHL Numerical Reasoning test typically contains 18 to 25 questions in 17 to 25 minutes — giving you approximately 60 to 80 seconds per question. The exact timing depends on which variant your employer uses, but the per-question window is broadly consistent. Some employers use a 17-question version with a 17-minute limit; others use a 25-question test with a 25-minute limit.

Sixty to eighty seconds sounds manageable until you consider what it actually covers: reading the question, locating the relevant data in the table or chart, performing the calculation (often a two-step calculation), checking your answer, and selecting the correct option. Under genuine test conditions, this feels fast.

A useful time management strategy is the 90-second rule: if you have not made meaningful progress on a question within 90 seconds, make your best guess and move on. An unanswered question scores zero. A guessed answer has a 20% chance of being correct — and if you have eliminated even one wrong option, that rises to 25%. Never leave a question blank.

SHL provides an on-screen calculator for most online numerical reasoning tests. Use it — even for calculations that feel simple enough to do mentally. Mental arithmetic under time pressure and anxiety introduces errors that a calculator prevents entirely.

A common mistake is trying to calculate something like "what is 23.4% of 187,000?" in your head to save a few seconds. This almost always takes longer than using the calculator and introduces an error risk that simply does not exist if you type it in. The few seconds you save are not worth the accuracy risk.

If you are permitted to bring a physical calculator (some employers allow this for unsupervised tests — check your invitation email), bring one you are genuinely comfortable with. Do not experiment with a new device on test day. Familiarity matters.

One important exception: estimation before calculation. Before typing numbers into the calculator, do a quick mental estimation of what the answer should approximately be. This takes 2–3 seconds and serves as a sanity check. If your calculated answer is wildly different from your estimate, you have probably mis-read the data or miskeyed a number.

This is the single most impactful time-saving strategy for SHL Numerical Reasoning: always read the question before scanning the data table or chart. Most candidates do the opposite — they read the dataset first, try to absorb all the numbers, and then read the question. This forces them to re-read the table a second time with the specific question in mind.

When you read the question first, you know exactly what you are looking for before you look at the data. This means you can scan directly to the relevant row, column, or bar and extract the specific numbers you need — ignoring everything else. On a data table with 30 cells and only 2 relevant to your question, this can save 15–25 seconds per question, which adds up to 4–7 minutes across a 20-question test.

This technique is especially valuable when questions reference a shared dataset. If a set of four questions all use the same table, reading the question first for each one means you are never re-reading the whole table — you are making targeted extractions each time.

If there is one calculation type that defines SHL Numerical Reasoning tests, it is percentage change. This question type appears in some form in the majority of SHL numerical tests, and candidates who have not memorised the formula reliably lose time working it out from first principles under pressure.

The percentage change formula is: ((New Value − Old Value) ÷ Old Value) × 100. That is all. Memorise it. Write it on a sticky note and put it next to your monitor for your practice sessions until it is automatic. On test day, you should be able to apply it without thinking.

Common trap: when a question asks for the percentage increase from Year 1 to Year 3, candidates often use Year 2 as the base rather than Year 1. Always use the starting year — the denominator in the formula — as the "old value," regardless of what looks most intuitive from the chart.

SHL Numerical Reasoning questions have five answer options. On most questions, three of those five options are clearly wrong — they are either the wrong order of magnitude, the wrong direction (a decrease instead of an increase), or obviously implausible given the data. Spotting and eliminating these three takes just a few seconds and dramatically improves your odds.

If you can eliminate three options, you are choosing between two. If your calculation points to one of those two, you can be confident. If you are running out of time and have to guess between the remaining two, you have a 50% chance — far better than a 20% guess across all five.

Elimination is also your best strategy when you are running out of time. If you have 15 seconds left on a question, a quick estimation that eliminates three options and lets you guess between two is far better than leaving the question blank. The test does not penalise wrong answers — only blank ones score zero.

Timed practice is not just about getting faster — it is about building the psychological tolerance for the time pressure that the real test creates. Candidates who only practise casually, without a timer and without consequence, are not practising the test. They are practising arithmetic. These are different skills.

Real-time practice means: setting a strict per-question timer (65–80 seconds), using an interface similar to the real SHL platform, having to commit to an answer before moving on, and reviewing your results at the end without being able to go back and change answers. This simulates the test experience closely enough to make the real thing feel familiar rather than shocking.

• Week 1: Untimed practice to learn question formats and identify your weak calculation types.
• Week 2: Timed practice at 90 seconds per question — slightly relaxed to build speed without panic.
• Week 3 onwards: Full timed practice tests at 65–75 seconds per question, simulating real test conditions. Review every wrong answer immediately after each test.

The most common preparation mistake — after not practising at all — is completing practice tests without properly reviewing incorrect answers. Most candidates look at their score, feel satisfied or disappointed, and move on. This wastes the most valuable learning opportunity in the entire preparation process.

When you get a question wrong on a practice test, you need to understand three things: (1) what the correct answer is, (2) why your answer was wrong — specifically which step you mis-executed, and (3) which underlying skill gap that reveals — a misapplied formula, a misread data unit, a failure to check the question before the table.

Each wrong answer is a precise diagnostic. A reverse percentage error tells you to drill reverse percentages. A units error (confusing £000s with £) tells you to build a habit of checking labels before calculating. A time management error (ran out of time and guessed randomly) tells you to practise elimination more systematically. Without reviewing wrong answers, you repeat the same mistakes across every test.