N-Back Test - The Working Memory Trainer - Free Online

About the N-Back Test (2-Back)

The N-Back is the most researched brain-training task in cognitive science. Squares light up one after another on a grid, and your job is to press MATCH whenever the current position is the same as it was exactly two steps ago. Sounds simple; feels like juggling. You must constantly update a two-item buffer in your head while ignoring everything else — pure working-memory load. This is the task behind the famous "can brain training raise IQ?" studies.

How to play

Press Start. Squares light up one at a time, every couple of seconds.
Hold the last two positions in your head at all times.
When the lit square matches the position from 2 steps ago, press MATCH.
Hits score points; false alarms cost them. 24 steps per round.

What it measures

This test focuses on your working-memory updating and sustained attention. Catching most matches with zero false alarms is a strong 2-back performance. Researchers use accuracy above ~80% as the cue to move to 3-back.

Why play it

The most studied working-memory training paradigm in cognitive science
Trains continuous updating — the skill behind mental arithmetic and multitasking
Short rounds with a precise score, ideal for daily tracking

Did you know?

The N-Back task was introduced by Wayne Kirchner in 1958.
A 2008 study by Jaeggi and colleagues linked dual N-Back training to fluid-intelligence gains — still debated, heavily cited.
Working-memory load in N-Back shows up clearly in prefrontal-cortex activity in brain-imaging studies.
The "dual" version adds simultaneous audio letters — same rules, two streams at once.

What counts as a good score?

Ranges below are drawn from published research — sources are listed at the end of the page.

Level	Accuracy at 2-back	What it means
Developing	Under 60%	Usually still learning the task itself. Almost everyone improves substantially over the first few sessions.
Average	60–75%	Solid 2-back performance once the task structure is familiar.
Above average	75–90%	Reliable updating — you are letting old items go cleanly instead of accumulating clutter.
Exceptional	Over 90%	Sustained near-ceiling accuracy. At this point 3-back is the more informative test.

The n-back is a standard laboratory measure of working-memory updating. Accuracy varies widely with stimulus timing and instructions, so these bands describe this implementation rather than serving as clinical norms.

How to improve your n-back score

1.Keep a rolling window of two

Hold exactly two items and actively discard the third. The failure mode is hoarding, not forgetting.

2.Say the pair out loud

Sub-vocalising "K, then M" makes the update explicit and much harder to lose under time pressure.

3.Do not count positions

Tracking how many items have passed adds load and breaks down fast. Compare against the item two back, not against a counter.

4.Accept the misses

A missed item costs one trial. Stopping to work out what you missed costs the next three.

5.Short sessions

N-back is exhausting and accuracy falls off sharply with fatigue. Ten focused minutes beats thirty distracted ones.

The claim that made n-back famous

In 2008 Susanne Jaeggi and colleagues reported in PNAS that training on a dual n-back task improved fluid intelligence — and that the gain scaled with training time. Fluid intelligence had been considered largely fixed in adults, so the paper was extraordinary, and it launched an entire brain-training industry.

The follow-up literature has been much less kind. Several well-powered replication attempts found gains confined to the trained task and to closely similar tasks, with little transfer to general reasoning. A 2016 meta-analysis by Melby-Lervåg and colleagues found near-transfer effects but no reliable far transfer.

So is it worth doing?

The honest position is that n-back reliably makes you better at n-back and at tasks that resemble it, and that the evidence for it making you generally smarter is weak. That is a much less exciting claim than the one on the box, but it is where the evidence currently sits — and it is still a demanding, well-constructed measure of working-memory updating, which is what this page offers it as.

Frequently Asked

What is the N-Back test?

A working-memory task where stimuli appear in sequence and you respond whenever the current one matches the stimulus from N steps earlier. Ours is a 2-back with grid positions: press MATCH when the lit square repeats from two steps ago.

Does n-back training increase IQ?

Probably not in any general sense. The 2008 Jaeggi finding was influential but has not replicated robustly, and meta-analyses find gains largely confined to the trained task. Expect to get better at n-back, not at everything.

What is a good N-Back score?

On our 24-step 2-back, catching nearly all matches with no false alarms is strong. In research settings, consistent accuracy above about 80% is the signal to progress to 3-back.

What is dual n-back?

A harder variant tracking two independent streams — typically position and sound — at once. It was the version used in the original training studies. This page implements single 2-back, which is the standard laboratory measure.

Why is 2-back so much harder than it looks?

Because it requires continuous updating rather than storage: every new item means promoting one and discarding another, under time pressure, with no pause. Two items sounds trivial until you have to maintain the operation for sixty trials.

N-Back Test - The Working Memory Trainer