You can test short-term memory with simple exercises that measure how many items you can hold in mind and recall after a brief delay. The classic benchmark is around seven items, plus or minus two, though more recent work suggests the true limit for most people is closer to four or five chunks of information when you can’t group items together. Whether you’re curious about your own cognitive baseline or concerned about a loved one, several well-validated tests range from quick at-home checks to formal clinical assessments.
The Digit Span Test
The digit span test is one of the most widely used short-term memory measures, and you can do a basic version at home with nothing more than a partner and a list of random numbers. One person reads a sequence of digits aloud at a pace of one number per second. The other person repeats them back in the exact same order. You start with a short sequence (two or three digits) and add one digit each round. When the listener fails both attempts at a given length, the test stops. The longest sequence recalled correctly is the “forward span.”
A backward version adds difficulty: after hearing the numbers, you say them in reverse order. If the reader says “4-9-2,” you’d respond “2-9-4.” This version taxes not just storage but mental manipulation, pushing into what researchers call working memory. Most adults can handle about six to seven digits forward and four to five backward. Scoring in formal settings awards one point if you get at least one of two trials correct at a given length, and two points if you nail both trials.
The Word Recall Test
Word-list recall is the backbone of most clinical memory screens. The format is straightforward: someone reads you a short list of unrelated words (typically five), and you repeat as many as you can. This is done twice to give your brain two chances to encode the list. That covers immediate recall.
The more revealing part comes about five minutes later, after you’ve been doing something else entirely. You’re asked to recall the same words without hearing them again. This delayed free recall is what clinicians pay the most attention to, because it separates people who encoded the information from those who didn’t. If you can’t recall a word freely, you might be given a category hint (“it was an animal”), and if that doesn’t help, a multiple-choice recognition prompt. Freely recalling a word scores the highest; needing a hint scores less; recognizing it from a list scores the least. The pattern of how you fail matters: someone who recognizes a word after a cue probably stored it but had trouble retrieving it, while someone who can’t even recognize it may not have encoded it at all.
The Corsi Block-Tapping Test
Not all short-term memory is verbal. The Corsi block-tapping test measures spatial short-term memory, your ability to remember locations in sequence. In the original version, nine small blocks sit in a scattered arrangement on a board. The examiner taps a sequence of blocks, and you replicate the pattern in the same order. Like the digit span, sequences start short (two or three blocks) and grow longer until you fail. Most people top out around five to six blocks, which is slightly lower than the verbal digit span. Digital versions of this test are widely available online and work the same way, highlighting squares on a screen instead of using physical blocks.
The N-Back Test
The n-back test is a continuous, self-paced challenge that pushes working memory harder than simple span tests. You see (or hear) a stream of items, one at a time, and your job is to signal whenever the current item matches the one presented “n” steps earlier. In a 1-back version, you’re just comparing each item to the one right before it. In a 2-back, you’re comparing it to the item two steps ago, which means you have to constantly update what you’re holding in mind. A 3-back raises the difficulty further.
What makes the n-back useful is that it directly scales cognitive load. Your accuracy at each level tells you how well your working memory system handles increasing demands. Many free apps and browser-based tools offer n-back training, and they typically adjust the difficulty in real time to keep you working at your limit. This isn’t just a one-time test; some people use it as a regular exercise, though the evidence on whether n-back training transfers to general cognitive improvement is mixed.
Working Memory vs. Simple Storage
It helps to understand that “short-term memory” actually covers two related but distinct abilities. Simple short-term memory is passive storage: holding a phone number in your head long enough to type it. Working memory adds active manipulation: rearranging that phone number into alphabetical order by the letter each digit starts with, for instance. Tests that only ask you to repeat information back (forward digit span, basic word recall) measure storage. Tests that require you to transform information (backward digit span, n-back) measure working memory, which draws on attention and processing speed in addition to raw storage capacity.
This distinction matters because someone can score normally on simple span tasks while struggling with complex ones, or vice versa. If you’re testing yourself or someone else, using both types gives a fuller picture.
Self-Administered Screening Tools
If your concern goes beyond curiosity and you want a more structured check, the Self-Administered Gerocognitive Exam (SAGE) is a pen-and-paper test developed at Ohio State University that you can take at home in about 15 minutes. It covers memory alongside orientation, language, and reasoning. In validation studies, SAGE detected mild cognitive impairment with 79 percent sensitivity and a false positive rate of just 5 percent. Its questions are deliberately harder than those on older screening tools like the Mini-Mental State Examination, specifically to catch very mild impairments that easier tests miss. You can download SAGE for free from Ohio State’s website and bring your completed test to a doctor for scoring and interpretation.
Online memory tests from research institutions can also provide a rough benchmark, but keep in mind that your testing environment matters. Distractions, fatigue, stress, and even the time of day can shift your scores significantly. A single poor result on any self-test is not a diagnosis.
What Poor Results Can Look Like
Occasional forgetfulness is normal at every age. You walk into a room and forget why, or you blank on a name for a few seconds before it comes back. That’s not a red flag. The patterns worth paying attention to are different in character: repeatedly asking the same question within a conversation, consistently struggling with tasks that used to be routine (like following a familiar recipe), misplacing objects in unusual places (a wallet in the refrigerator), or getting disoriented in a neighborhood you know well. Mixing up words in odd ways, like saying “bed” when you mean “table,” is another signal that goes beyond garden-variety forgetfulness.
On formal tests, the specific warning signs include an inability to recall any words after a five-minute delay even with category cues, or a forward digit span consistently below four, which falls well outside the normal range. A single bad day doesn’t mean much, but a pattern of declining performance over repeated testing, or scores that surprise you given your baseline, is worth discussing with a neurologist or neuropsychologist who can run a comprehensive evaluation.
How to Get the Most Accurate Results
If you’re testing yourself or someone else informally, a few practical steps improve reliability. Test in a quiet room with no background noise or screens competing for attention. Run the test at the same time of day if you plan to compare scores over weeks or months. Avoid testing right after poor sleep or alcohol use, both of which temporarily suppress short-term memory performance. And use more than one type of test: combine a verbal task (digit span or word recall) with a spatial one (Corsi blocks or an online equivalent) to capture different dimensions of memory.
For a true clinical picture, formal neuropsychological testing remains the gold standard. These sessions typically run one to three hours and assess memory alongside attention, language, executive function, and processing speed, giving a detailed cognitive profile rather than a single number.