What to Do in the Last 10 Minutes of an FLK Paper

Ten minutes left. The clock in the corner of the screen has gone from a background hum to the loudest thing in the room. You've got four questions flagged, one you never answered at all because you skipped it in a panic around question 40, and a nagging feeling that you misread a Business Law fact pattern somewhere in the first hour. What you do with those 600 seconds can genuinely be the difference between a pass and a resit.

Most SQE exam preparation focuses on the first 140 minutes — pacing, elimination, reading the stem before the options. Fair enough; that's where the bulk of the marks live. But the endgame is a separate skill, and almost nobody practises it. Candidates rehearse full mocks and then let the last stretch dissolve into vague panic-scrolling. That's a waste of the most marginally productive minutes in the whole assessment.

Why the final ten minutes are worth more than the first ten

Remember the shape of the thing. Each SQE1 assessment — FLK1 and FLK2 alike — is 180 single-best-answer questions, split across two sessions of 2 hours 33 minutes on the same day. That's 90 questions per session in 153 minutes. Do the arithmetic and you get roughly 1 minute 42 seconds per question, which sounds generous until you meet a three-paragraph Solicitors Accounts problem with a ledger entry buried in the fourth line.

Here's the thing about the last ten minutes: the marks available there are cheap. At the start of a session you're spending 102 seconds to earn one mark on a question you'd probably get right anyway. At the end, you might convert three near-misses into three marks in the same time — because the hard analytical work is already done and all that's left is a decision. Flagged questions are pre-loaded. You've read the facts, you've eliminated two options, you just couldn't commit. Coming back with a fresh head and a hard deadline forces the commitment.

And there's no negative marking. An unanswered question and a wrong answer score exactly the same: nothing. Which means a blank on your screen at the moment the session closes is a pure, avoidable loss. I've spoken to candidates who ran out of time with six unanswered questions on FLK2 and then agonised for weeks over whether that was the gap. With five options per question, six random guesses give you roughly one extra mark on average. Not life-changing. But on a borderline script, one mark is the whole ballgame.

The ten-minute protocol

Rehearse this until it's automatic. The point of a protocol is that it runs when your prefrontal cortex has gone offline.

Minutes 10 to 7: kill every blank

This comes first because it's the only step with a guaranteed positive expected value. The Pearson VUE interface gives you a review screen showing answered, unanswered and flagged items — familiarise yourself with exactly how it looks and behaves before exam day, because fumbling with navigation while the clock burns is a special kind of hell. (The on-screen tutorial at the start of the session is worth genuinely reading, not clicking through.)

For each blank: give yourself ten seconds. Skim the last line of the stem — the actual question — and the options. If one leaps out, take it. If nothing leaps out, pick your default letter and move on. Do not fall in love with a Land Law easement problem at this stage. You are harvesting, not farming.

Minutes 7 to 2:30: triage the flags, cheapest first

Not all flagged questions are equal, and this is where most candidates go wrong. The instinct is to return to the hardest one — the Trusts question about whether a purported declaration of trust over an unascertained part of a shareholding was sufficiently certain — because it feels unfinished. Resist. That question defeated you when you had time and calm. It will defeat you again.

Instead, sort mentally into two buckets:

• Two-horse races. You've genuinely narrowed it to options B and D and you're 60/40. These are worth 45–60 seconds each. Re-read the question stem — not the facts, the question — because in a majority of two-horse races the tie-break is hiding in what's actually being asked. "Which is the best advice to the client?" behaves very differently from "Which statement best explains why the claim will fail?"
• Fog. Three or more options still live, no clear line of attack. These get fifteen seconds and a decision. Nothing more.

Do the two-horse races first, in the order you meet them going down the review screen. Don't hunt around for the "best" one to fix; the hunting costs more than it saves.

The last ten minutes are not for solving problems. They are for converting almost-decisions into decisions, and blanks into guesses. Solve nothing new.

Minutes 2:30 to 0:45: the completeness sweep

One pass down the review screen. Eyes only on the answered/unanswered column. You are looking for a single thing: a gap. It happens more than you'd think — a candidate clicks an option, the click doesn't register properly, and they navigate onward none the wiser. Thirty seconds of scrolling insures against that.

If you find a blank at 1:40 remaining, that's fine. Guess it. That's what the sweep is for.

Guessing properly (yes, there's a right way)

Blind guessing on a five-option single-best-answer question gives you a 20% shot. Eliminating one option takes you to 25%. Eliminating two takes you to 33%. That difference compounds meaningfully across a paper, and elimination is fast when you know what to look for.

Under extreme time pressure, these heuristics earn their keep on SQE1:

1. Bin the absolutes. Options containing "always", "never", "in all circumstances", "must in every case" are wrong more often than they're right, because English law is a discipline of exceptions. Not a rule — a lean.
2. Bin the answers that are legally true but don't answer the question. The SRA's question writers are fond of the correct-but-irrelevant distractor. If the stem asks about the solicitor's obligation under the Code of Conduct and an option gives you an impeccable statement of contract law, it's not your answer.
3. Watch for the option that ignores a fact you were deliberately given. If the fact pattern told you the buyer is a company, or that the deceased died intestate leaving no spouse, that detail is load-bearing. An option that would read identically without it is usually the trap.
4. Have a default letter. Pick one — C is as good as any — and use it for anything you cannot eliminate at all. Not because C is more likely to be right, but because a pre-made decision costs zero seconds and stops you dithering. Dithering is the real enemy.

Should you change an answer you've already given?

Cautiously, and only for a reason you can name. The rule I'd give any candidate: change it if you can articulate what you got wrong, don't change it if you just feel uneasy. "I misread — the client is the seller, not the buyer" is a reason. "Hmm, D looks more like a real answer now" is not; that's exam fatigue talking, and it costs people marks every sitting.

The dangerous window is the final ninety seconds, when adrenaline makes every previously-confident answer look suspicious. This is why the protocol above deliberately gives you nothing to do at the very end. Idle hands click things.

A worked example: eleven minutes on FLK2

Take a concrete case. Priya is in the afternoon session of FLK2. At the 11-minute mark her review screen shows:

• Two unanswered questions (Q34, a Solicitors Accounts ledger problem she skipped deliberately; Q71, which she simply never reached the first time through).
• Five flagged: Q12 (Land Law, covenant enforceability), Q29 (Wills, whether a gift adeemed), Q46 (Criminal Practice, bail conditions), Q58 (Property Practice, pre-contract enquiries), Q83 (Trusts, breach of trust remedies).

What she actually does:

11:00 – 08:20. Q34 and Q71 get answered. Q34's ledger arithmetic is beyond rescue in the time available, so she looks at whether the options describe client account or business account entries, eliminates the two that would put client money in the wrong place, and picks from the remaining three. Q71 she reads properly — it turns out to be a straightforward question about the effect of a caveat and she gets it in forty seconds. That's a mark she would have thrown away entirely.

08:20 – 03:00. Q29 and Q58 were both two-horse races, so they go first. On Q29 she re-reads the question line and realises it asked what the residuary beneficiary receives, not whether the specific gift failed — that reframing decides it in twenty seconds. Q58 stays a coin flip; she commits and moves on. Q46 and Q12 get fifteen seconds each and a decision. Q83, the Trusts monster that ate four minutes earlier in the session, she does not reopen. She leaves her original answer standing.

03:00 – 01:00. Completeness sweep. All 90 answered. She finds nothing, which is exactly what you want a sweep to find.

Final minute. Hands off the mouse.

Best guess at the net effect: one certain mark on Q71, one probable on Q29, and improved odds on three others. Call it two to three marks. On a paper where the pass threshold and your score might be a hair apart, that's not trivial — and every one of those marks came from process, not knowledge.

Building the endgame into your revision

You cannot execute a protocol you've never run. Which means the last ten minutes have to be part of your SQE revision, not an afterthought you improvise on the day.

Practical ways to build it:

• Do timed blocks that end abruptly. Set a timer for a 90-question block at full exam length and let it cut you off. Sitting there with two unanswered questions when the timer dies is an unpleasant but instructive experience. You only need it once.
• Deliberately over-flag in practice. Flag anything you're less than 80% sure of. You'll finish a mock with fifteen flags and no time to clear them, which teaches triage in a way that reading about triage never will.
• Track what your flags are actually worth. After a mock, check how many flagged questions you got right on your first instinct versus after revisiting. Some candidates discover their first instinct is better than their second thoughts. That's a strategically useful thing to know about yourself before exam day.
• Practise the guess. Force yourself to answer a question in ten seconds using elimination alone. It's a distinct skill, and it's the one you'll be relying on at 9:30 remaining.

This is the case for volume practice under realistic conditions rather than passive re-reading of notes. Working through a large, well-tagged bank — the Ant Law SQE Question Bank covers 14,000-plus single-best-answer questions mapped across all thirteen FLK1 and FLK2 subjects, with 90-question timed mocks that mirror the SRA's format and timing ratio — is how the endgame protocol stops being a list on a blog and starts being muscle memory. The wrong-answer book is where the real learning happens; the mock timer is where the nerve gets built.

One thing to sort out before the day itself

Know how the review screen works. Know how flagging works. Know whether you can navigate freely within the session. These are small operational details, and every year candidates lose minutes to them because they assumed it would be obvious. The SRA publishes guidance and a familiarisation walkthrough — go and look at sqe.sra.org.uk and spend twenty minutes with it. Twenty minutes now, against four minutes of confusion in the exam, is an excellent trade.

Keeping it in proportion

None of this substitutes for knowing your law. First-attempt pass rates on FLK1 and FLK2 have historically hovered around the halfway mark — check the SRA's most recent published statistics for the current position — and the candidates who fall short usually do so because of thin coverage across thirteen subjects, not because of ten minutes of poor triage. Exam technique is a multiplier on knowledge, never a replacement.

But it is a multiplier. And when you consider what sits around this exam — the qualifying degree or equivalent, two years of qualifying work experience (QWE) to log and have confirmed, the SRA character and suitability assessment, SQE2's five practical skills waiting on the other side — it seems daft to leave marks on the table for want of a plan you could rehearse in an afternoon. The route to solicitor qualification in England and Wales is long enough without adding an avoidable resit to it.

So: write the protocol on an index card. Ten to seven, blanks. Seven to two-thirty, flags cheapest first. Two-thirty to forty-five seconds, sweep. Then hands off. Run it in your next three timed mocks until you stop needing the card.

If you want somewhere to run those mocks properly — full-length, timed, with the flag-and-review rhythm baked in — try a timed 90-question block on antlaw.ai and pay attention to what you do with your final ten minutes. That's the bit nobody else is practising.