Ability is one of the five elements of Success Profiles, and for many Civil Service roles it is assessed before a human ever reads your application. That ordering matters: a strong personal statement is worth nothing if you have already been screened out by a test you attempted on your phone on a train.
The tests are not designed to be beaten, but they reward familiarity heavily, and most candidates arrive at them cold. This guide covers what each one involves and where preparation actually pays off.
The main tests you will meet
Exact names and formats change, and different campaigns use different combinations, so treat your invitation email as the authority. In recent years the common ones have been:
Civil Service Judgement Test (CSJT). A situational judgement test. You are given a realistic workplace scenario and asked to rate the effectiveness of several possible responses, or pick most and least effective. Used very widely, including at EO and HEO level and on the Fast Stream in a work based scenarios format.
Civil Service Verbal Test (CSVT). A passage of text with statements to be judged true, false or cannot say based only on the passage.
Civil Service Numerical Test (CSNT). Data in tables and charts with questions requiring interpretation and calculation, under time pressure.
Some technical and analytical roles add further specific assessments, and some campaigns include a Behaviours or motivation questionnaire alongside the ability tests.
The judgement test is the one to understand properly
The CSJT is where preparation is most misunderstood. Candidates try to revise for it as though it were knowledge based, then conclude it cannot be prepared for at all. Both are wrong.
You cannot learn the answers, but you can understand the value set being tested, and that materially changes how you rate options. The Civil Service is consistently looking for responses that:
- Consult the right people before acting, rather than either escalating everything or ploughing on alone
- Deal with the issue directly rather than avoiding it or going around a person
- Consider the wider impact, including on other teams, the public, and the department's reputation
- Are honest and proportionate, neither concealing a problem nor escalating a minor one
- Follow process where process exists, while showing judgement about when it needs flexing
Two patterns are worth knowing because they recur. Options that involve doing nothing and hoping, or that involve going straight to a senior manager over a colleague's head as a first move, are usually rated poorly. Options that involve speaking to the person concerned first, gathering the facts, and then escalating if needed are usually rated well. This maps closely to what makes a good situational interview answer, which is not coincidental: both assess the same judgement.
The most useful preparation is to read the Behaviours framework before sitting the test. The judgement being assessed is the judgement described in those descriptors, so the framework is effectively a study guide for the CSJT. The Behaviours breakdown is the place to start.
Verbal reasoning: the "cannot say" trap
The verbal test looks easy and catches people out on one specific mechanic. You must judge each statement using only the passage, not using what you know to be true in the world.
The single biggest source of lost marks is answering "true" for a statement that is plausible and probably correct in reality but not actually stated or logically entailed in the text. The correct answer in those cases is "cannot say", and getting comfortable with that feels counterintuitive.
The other frequent error is over-reading qualifiers. Words like "all", "most", "only", "may" and "usually" carry the weight in these questions. A passage saying most regions reported an increase does not support a statement that all regions did.
Practical technique: read the statement first, then scan the passage for the relevant sentence, rather than reading the whole passage carefully in advance. Time is the constraint, not comprehension.
Numerical: familiarity beats mathematics
The mathematics is not advanced. Percentages, ratios, percentage change, reading multi-column tables and interpreting charts covers most of it. What makes the test hard is time pressure combined with awkwardly presented data.
Where preparation pays:
- Know the common calculations cold, especially percentage change and reverse percentages, so you are not deriving the method under pressure
- Practise reading the question before the table. Most tables contain more data than any one question needs, and reading everything first wastes the time you need
- Check the units and the footnotes. Figures in thousands or millions, and footnotes excluding a category, are where careless errors come from
- Learn to abandon a question. All questions typically score equally, so spending ninety seconds rescuing one is a poor trade against answering two easier ones
Use a calculator where permitted and have paper. The general points in the psychometric test guide apply directly here.
The practical mistakes that cost places
Attempting tests in an unsuitable setting. A meaningful number of candidates lose out because they sat a timed test on a phone, in a noisy environment, or in a fifteen minute gap. Block out the time properly, use a laptop or desktop, close everything else, and have water, paper, pen and a calculator to hand.
Not doing the practice tests provided. The Civil Service usually supplies official practice materials with your invitation. These are the closest available match to the real format, and they are free. Do them, in full, timed, before the real attempt.
Treating the deadline casually. Test invitations carry a window, often around a week. Leaving it to the final evening removes any ability to reschedule around a technical problem, and technical problems do happen.
Ignoring accessibility adjustments. If you have a disability or condition affecting how you take tests, adjustments including extra time are available, and you should request them at application stage rather than after. This is a normal and expected part of the process, not a favour, and there is more on the broader topic in the interview adjustments guide.
Assuming a poor result closes the door permanently. Retake policies vary by campaign and by test, with some allowing another attempt after a set period. Check the specific campaign rather than assuming either way.
How the tests fit the wider process
Ability tests are a screen, not the assessment. Passing them gets your written evidence read; it does not build a case for you. The case is made in your Statement of Suitability and your Behaviour evidence, and then at interview through Behaviours and Strengths.
That means the right allocation of preparation time is: enough on the tests to be familiar and unflustered, then the bulk of your effort on the evidence that humans will score. Understanding how that scoring works is where the largest gains are, because the difference between a 4 and a 6 is mostly a matter of what you chose to write down.
For the process end to end, including how Sift and Interview fit together, see the Civil Service interview guide, and if you are applying to the graduate scheme, the Fast Stream guide covers the additional stages.