UN written tests and technical assessments: what to expect and how to prepare

Between shortlisting and the panel interview, many UN and IO vacancies add a stage candidates rarely hear about until they're in it: a written test or technical assessment, sat alone, against the clock, before anyone speaks to you. It filters harder than the interview does, because there's no panel to read your intent — only what you produced in the time given.

This guide covers what these assessments actually look like across the system, how they're scored, and how to prepare for one without knowing the exact questions in advance. For where this stage sits in the wider process, see the recruitment timeline guide; for the interview that usually follows it, see how panels actually score you.

Why organizations add a written stage

A competency-based interview tests how well you talk about your work. A written or technical assessment tests whether you can actually do it, under realistic constraints, with no coaching and no second draft. Organizations add this stage specifically for roles where the job itself is writing, analysis or a technical skill — drafting, policy analysis, finance, data, translation, IT, procurement — because a strong interview performance doesn't reliably predict a strong first memo.

It also does real filtering work before the more expensive panel stage: a hiring manager who has to interview five candidates instead of fifteen has a written or technical result to thank for the shorter list.

The formats you'll actually meet

"Written test" covers several distinct formats. Which one you get depends almost entirely on the job family:

• Drafting exercise. Produce a memo, briefing note, press statement or talking points from a prompt and sometimes a short set of background documents, under a fixed word count and time limit. Common for policy, communications, public information and programme roles.
• Case study / in-tray exercise. A simulated inbox or scenario — conflicting instructions, incomplete information, a deadline — asking you to prioritise, decide and justify. Common for management and coordination posts.
• Technical knowledge test. Structured questions against a specific body of knowledge: financial regulations for finance roles, procurement rules for supply chain, a coding or systems exercise for IT, statistical methods for data and M&E posts.
• Language or translation test. A timed translation, précis or editing exercise, standard for language posts and frequently added as a secondary filter wherever a working language beyond English is required.
• Timed online test platform. Multiple-choice or short-answer verbal, numerical and abstract-reasoning batteries, often outsourced to a third-party testing platform, most common at the EU institutions' computer-based testing stage and increasingly used by larger UN agencies at the longlist stage.

Some processes combine two of these — a short knowledge test followed by a drafting exercise in the same sitting — so read the invitation email closely for format, timing and whether the platform allows reference material.

Where it sits in the process

Most commonly the written or technical stage runs after shortlisting and before the panel interview, filtering the shortlist down to a smaller group who get interviewed. Some organizations instead run it alongside or immediately after the interview, using it to confirm a provisional ranking rather than to filter. A minority fold a short knowledge test into the online application itself, at the very top of the funnel. The invitation you receive will state where in the sequence you are — if it doesn't, it's a reasonable question to ask the HR contact listed on the vacancy.

What graders are actually scoring

Written assessments are usually scored against a rubric tied directly to the competencies and technical criteria in the vacancy announcement, not against literary quality. In practice, graders are looking for:

• Structure over style. A clearly organised answer — the point stated early, reasoning that follows logically, a usable conclusion or recommendation — scores better than well-turned prose that buries the point.
• Relevance to the actual prompt. Answering the question that was asked, including every sub-part, beats a more impressive answer to a related but different question.
• Evidence of the specific skill being tested. A drafting exercise is scored on drafting — tone, accuracy, audience-appropriate register — not on the underlying subject expertise alone.
• Fit within the constraints. Word counts and time limits are usually part of the test itself: a technically strong answer that ignores the stated word limit signals an inability to work within real operational constraints.

How to prepare without knowing the questions

You can't rehearse the exact prompt, but the format and the subject area are both knowable in advance:

• Read the organization's own output. If you're being tested on a drafting exercise, read several recent press statements, briefing notes or reports the actual unit has published. Register and tone matter, and they're freely available.
• Practise against a clock, not just against a topic. The binding constraint in most of these tests is time, not knowledge. Write a full response to a plausible prompt in the stated time limit at least once before the real thing.
• Know the criteria the vacancy announcement already gave you. The competencies and technical requirements listed in the posting are usually the rubric in disguise — structure your practice answers to visibly demonstrate each one.
• Check what's allowed. Some tests permit open reference material (regulations, a laptop, internet access); others are closed-book. Confirm this from the invitation before test day so you aren't caught preparing the wrong way.
• Protect the sitting itself. Online timed tests are usually a single attempt with no pause once started. Clear the block of time, confirm your connection, and read every instruction screen before starting the clock.

The mistakes that actually cost candidates

• Running out of time on the first section and leaving a later, equally weighted section blank or rushed.
• Writing the answer you wanted to give instead of the one the prompt asked for — common when a candidate has a prepared answer from research and forces it onto a slightly different question.
• Ignoring the stated audience. A briefing note for a senior manager and a public-facing statement are scored on different registers; using the wrong one reads as a real skill gap, not a stylistic choice.
• Treating a technical knowledge test as open-ended when it's closed-book, or the reverse — assuming no references are allowed and wasting preparation time memorising material a real assessment would have let you look up.

Once you know the format and what's actually being scored, this stage stops being a mystery and becomes something you can prepare for like any other part of the process. Browse current vacancies across the board, or go back to the complete field guide for how the rest of the pipeline is assessed.