PwC Online Assessment 2026: SHL Tests, Format & Preparation Guide

PwC UK hires several thousand graduates every year across six business lines: Audit & Assurance, Tax, Deals, Consulting, Technology, and Private Clients & Entrepreneurs. The recruitment funnel is structured in five stages, and the online aptitude assessment is Stage 2 — the first objective cognitive filter after the initial application.

Clearing Stage 2 is binary: you either pass the cut-off and progress, or you are screened out automatically. No other part of your application compensates for a low aptitude score at this stage. That makes the online tests the single highest-leverage preparation target in the entire PwC process.

This guide covers Stage 2 in full detail. If you are preparing for Stage 4, our separate PwC Assessment Centre guide covers the group exercise, written task, and partner interview in depth.

PwC's online assessment uses SHL's TalentCentral platform and is delivered remotely via your browser. The battery comprises three cognitive ability tests — numerical, verbal, and logical reasoning — plus a Situational Judgement Test (SJT). All four components are completed in a single session.

Logistics Before You Start

• PwC typically gives you five to seven days to complete the assessment from when the invitation link is sent — do not leave it until day six.
• The tests must be completed in one uninterrupted session; you cannot save progress and return.
• A calculator is permitted for the numerical section — have one within reach before you begin.
• The platform uses remote proctoring: tab-switching and unusual cursor behaviour are flagged automatically.
• Use a laptop or desktop. Taking aptitude tests on a phone or tablet significantly increases error rates due to small data tables and charts.

The numerical reasoning test is the most heavily weighted component of PwC's online assessment. Each question presents a data set — a table, bar chart, line graph, pie chart, or combination — followed by a multiple-choice question requiring a calculation or interpretation. The data always has a business or financial context: revenue breakdowns, market share comparisons, headcount statistics, cost allocations.

What makes SHL numerical questions demanding is not the underlying arithmetic — it is the multi-step structure combined with time pressure. A typical question requires you to extract two or three values from a chart, apply a formula (percentage change, ratio, margin), and select the correct answer from five plausible options. Each step is individually straightforward; doing all of them accurately in 75 seconds is where preparation makes the difference.

The Six Most Common Question Types

For a full breakdown of numerical reasoning question types, worked examples, and timed drills, see our SHL Numerical Reasoning guide.

The verbal reasoning test presents short passages — typically 100–250 words — drawn from business, management, or finance writing. After each passage you are given a statement and must decide whether it is True (follows directly from the passage), False (contradicts the passage), or Cannot Say (cannot be determined from the passage alone).

With roughly 40–60 seconds per question including reading time, pace is a genuine constraint. You cannot afford to read each passage twice. The skill the test measures is not vocabulary or general comprehension — it is the ability to evaluate the logical relationship between a passage and a statement, quickly and under pressure.

The Cannot Say Trap — the Most Important Rule in Verbal Reasoning

The most common and costly error on verbal reasoning tests is answering based on external knowledge rather than the passage. Even if you know a statement to be true in the real world, if the passage does not directly support it, the answer is Cannot Say. This feels counterintuitive but is absolute: the passage is the only valid evidence source.

Speed-Reading the Passage Efficiently

• Skim the statement first. Before reading the passage, glance at the statement to anchor your reading. You know what you are looking for, so you can read the passage with a specific filter.
• Hunt for quantifiers. Words like all, always, never, most, some, only are frequently the hinge on which a statement moves from True to Cannot Say. The passage may say "most clients" while the statement claims "all clients" — that difference makes it Cannot Say.
• Ignore distractors. Passages often contain information that is irrelevant to the specific statement. Do not let it slow you down.

The logical (also called inductive or diagrammatic) reasoning test removes domain knowledge entirely. Every question presents a sequence of abstract shapes, symbols, or grids — and your task is to identify the rule governing the sequence and select which option correctly continues or completes it.

Because no prior knowledge is relevant, this test measures pure pattern recognition and structured thinking speed. It is also the most trainable of the three cognitive tests: once you have internalised a systematic scanning framework, your accuracy improves dramatically and quickly — often within a week of deliberate practice.

The Five Pattern Variables

Every SHL logical reasoning question can be fully decoded by systematically scanning for changes across exactly five dimensions. Train yourself to scan them in this fixed order on every question:

Most questions involve two or three active variables simultaneously. By checking them in a fixed order rather than staring at the image hoping to spot a pattern, you eliminate the slow and inconsistent "holistic guess" approach that most unprepared candidates rely on.

For more on the seven recurring pattern rules and worked examples, see our SHL Inductive Reasoning guide.

The SJT presents realistic workplace scenarios — client requests, deadline conflicts, team disagreements, ethical dilemmas — and asks you to rate or rank a set of possible responses from most to least effective. Unlike the cognitive tests, the SJT is untimed, though completing it in 25–30 minutes is typical.

PwC's SJT is designed to assess alignment with the firm's five stated values: act with integrity, make a difference, care for others, work together, and reimagine the possible. Your responses are scored by comparing them to a benchmark provided by senior PwC professionals who completed the same scenarios.

Common Scenario Themes

• A junior colleague has made an error in a client-facing document with a short deadline — what do you do?
• A client asks you to present data in a technically accurate but potentially misleading way — how do you respond?
• Two managers give you contradictory instructions on the same project — how do you proceed?
• A team member is consistently underperforming but you have no direct authority over them — what is your approach?
• You notice a potential conflict of interest that your manager appears unaware of — do you raise it and how?

SHL uses norm-referenced scoring: your raw score on each test is compared against a benchmark population of previous test-takers at the same level (graduate, in this case). Your result is reported to PwC as a percentile rank — the proportion of the norm group you outperformed — and a sten score on a 1–10 scale.

PwC then applies a cut-off threshold — a minimum percentile below which candidates are automatically screened out. This cut-off is set by PwC independently for each business line and intake year, and it is not disclosed publicly. What is known is that the effective bar is higher for competitive lines (Consulting, Deals) than for others.

No Negative Marking

SHL assessments do not deduct marks for wrong answers. The implication is straightforward: always answer every question, even under time pressure. A considered guess has positive expected value; leaving a question blank scores zero with certainty. If you reach the end of the time with unanswered questions, mark every remaining question before the timer expires.

For a deeper explanation of how percentile ranks, sten scores, and norm groups work, see our SHL Test Scores Explained guide.

Structured practice over two to three weeks consistently produces meaningful percentile gains — typically 10–20 percentile points compared to unprepared candidates sitting the same test. The key mechanism is deliberate practice: reviewing every wrong answer immediately after each session to understand the specific technique or error involved, then drilling that gap until it is no longer a gap.

Generic exam revision (rereading notes, doing a couple of practice tests without review) produces almost no improvement. It is the review and targeted repetition that drives gains.

Eight High-Impact Preparation Rules

• Always use a calculator during numerical practice. Developing mental arithmetic habits that you then can't use under exam conditions trains the wrong skill. Practice exactly as you will perform.
• Never practise without a timer. The ability to perform under time pressure is a separate skill from raw ability. It only develops through repeated exposure to timed conditions.
• Review wrong answers immediately, not at the end of the week. Delayed review means you have forgotten the reasoning process that led to the error. Immediate review cements the correct technique in memory.
• For verbal reasoning, practise only the True/False/Cannot Say format. Other reading comprehension formats (multiple choice answers, open questions) do not transfer. Drill the three-option format exclusively.
• Name the pattern variables out loud for logical questions. Saying "Number: constant. Size: decreasing. Shade: alternating..." forces systematic scanning and prevents holistic guessing.
• Read PwC's values and Early Careers page before practising the SJT. Ten minutes of context preparation changes how you approach every scenario.
• Do your final full simulation the evening before the real test. Not the morning of — completing the test fresh the next morning, without an exhausted warm-up session immediately beforehand, produces better results.
• Test your technology at least 24 hours before your scheduled slot. Confirm your webcam works, your browser is compatible with SHL TalentCentral, and your internet connection is stable. Discovering a technical issue five minutes before you start is catastrophic for your focus.