The UCLA Loneliness Scale is the instrument behind a large share of what psychology actually knows about loneliness. When a study reports that loneliness predicts a health outcome, or that a particular group is lonelier than another, there is a good chance this scale, or its short-form relative, is what produced that number.
What it is, and where it came from
The scale was originally developed in 1978 and has been revised since, with the 1996 version by Russell becoming the standard now used across most research. [russell-1996-ucla] It consists of 20 statements, each rated on how often it applies, from “never” to “often”, producing a total score between 20 and 80. It was built and tested specifically to measure loneliness as its own construct, separate from related but distinct things like depression or social anxiety.
Reliability testing on the 1996 revision found internal consistency between .89 and .94, and test-retest reliability of .73 across a full year, both of which are strong figures for a psychological measure. [russell-1996-ucla] That combination, high consistency and reasonable stability over time, is a large part of why this particular scale became the field standard rather than one option among several competing ones.
What it measures, and what it deliberately does not
Every item on the scale asks about a felt experience: whether someone feels in tune with the people around them, feels part of a group of friends, feels isolated from others. None of it asks how many people someone lives with, how often they see friends, or how much time they spend alone. That is a deliberate design choice, because loneliness researchers distinguish the subjective feeling of connection from the objective fact of company, and the two do not track together as closely as intuition suggests.
That is also why the scale can produce a high score for someone who is rarely physically alone, and a low score for someone who spends most of their time by themselves. It is measuring the gap between the connection wanted and the connection felt, not the headcount in the room.
Illustrative summary of the scale's item content based on its published structure, not a measured statistic.
The short version people are more likely to actually take
Twenty items is a lot for a survey that is trying to measure a dozen other things at once, which is why a 3-item short form exists for exactly that use case. Hughes and colleagues developed and validated a version asking three simpler questions, how often someone lacks companionship, feels left out, and feels isolated, scored from 3 to 9, specifically for large population studies where the full scale was impractical. [hughes-2004-uls3] It correlates well with the full 20-item scale, which is what makes it usable as a stand-in, but it is a distinct, separately validated instrument rather than a shortened copy of the same test.
This distinction matters for anyone taking a “UCLA loneliness” quiz online: most quick checks, including the one on this site, use the 3-item short form rather than the full 20-item version, because that is the version built for a fast, single-purpose check rather than a multi-part research survey.
What a high score actually means
There is no single official cutoff that marks someone as clinically lonely on either version, and that is intentional: loneliness is not a diagnosis, and the scale was built to measure a continuum rather than to sort people into categories. A high score is still meaningful information. It says the gap between the connection someone has and the connection they want is currently large, which research consistently links to worse sleep, mood and physical health outcomes over time. That is worth acting on precisely because it is a feeling rather than a verdict: feelings respond to changed circumstances in a way a fixed diagnosis would not.
When to seek help
Speak to a doctor if a persistently high sense of loneliness is accompanying low mood, sleep problems or a loss of interest in things that used to matter, since those combined are worth a proper look regardless of what any one questionnaire says on its own. A single score, on any version of the scale, is a starting point for a conversation rather than a conclusion.
How MyFreud can help
Our quick loneliness check uses the validated 3-item short form described above, scored in the same tradition as the full research scale, as a fast, honest first read rather than a diagnosis. Our loneliness guide covers what drives it and what actually helps beyond the score itself.