At a glance
The Yale-Brown Obsessive Compulsive Scale has rated OCD severity since 1989. Its history, scoring bands, the Y-BOCS-II revision and its modern uses in trials and surgery.
A number between zero and forty decides, in most of the world's OCD clinics, how ill a person is. It comes out of a semi-structured interview drafted in New Haven in the late 1980s. It has outlasted the diagnostic manual it was written alongside, the rating fashions that followed it, and every serious attempt to replace it.
The Yale-Brown Obsessive Compulsive Scale — Y-BOCS to anyone who uses it — is a clinician-administered interview with ten scored items. Each runs from 0, no symptoms, to 4, extreme. The first five items rate obsessions, the last five rate compulsions, and the sum gives a severity score from 0 to 40, with separate subtotals for each half.
That is the whole architecture. It should not have been durable. It was built for drug trials, by psychiatrists who needed a common yardstick, and it has ended up in operating theatres, insurance letters and regulatory filings.
The Y-BOCS did not succeed because it captured obsession exactly. It succeeded because it made obsession portable.
Ten questions, five dimensions, one interviewer
Before a single point is awarded, the patient is walked through a symptom checklist. Goodman's team built one with more than fifty types of obsessions and compulsions arranged into fifteen broad categories: contamination, washing and cleaning, checking, symmetry and exactness, sexual and religious obsessions, hoarding, and the rest. The checklist exists because people leave things out. Someone will describe the handwashing and never mention the thought that God is counting, and an interviewer who does not ask will never find out.
The scoring half is narrower. Each of the ten items probes one of five things, and probes it twice — once for obsessions, once for compulsions. How much time do the symptoms take? How far do they interfere with work and relationships? How much distress do they cause? How hard does the person fight them? How much control does the person actually have?
Two patients with nothing in common, one terrified of contamination and one terrified of being a paedophile, get measured on the same five dimensions. Severity is the unit. Content is not.
Bands translate the total into ordinary language, and clinicians read them the way they read blood pressure ranges.
| Total score | Interpretation |
|---|---|
| 0 to 7 | Subclinical |
| 8 to 15 | Mild |
| 16 to 23 | Moderate |
| 24 to 31 | Severe |
| 32 to 40 | Extreme |
Those bands are a convention rather than a law of nature, attached to the instrument by the people who used it rather than the men who wrote it. They are now treated as fixed.
The interview is not quick. The original authors flagged the burden themselves, noting that a first assessment can run ninety minutes with follow-ups around half an hour. That cost is why shorter screens exist, and why the long version still gets used when the decision in front of the clinician is an expensive one.
The scales that were cut
The Yale OCD Research Clinic was founded in the mid-1980s by Wayne Goodman and Dennis Charney, and it had an advantage most psychiatry departments of the period lacked: enough severe cases to see patterns. The same clinic went on to run early trials of the drugs that became standard treatment, the serotonin reuptake inhibitors and the antipsychotics used as add-ons.
Goodman's team piloted considerably more than they published. Their early draft carried the ten core items plus a further set of independent scales covering insight, avoidance, indecisiveness, pathological responsibility, slowness, pathological doubt, and global severity and improvement. Six months of testing across more than forty patients convinced them to cut the extras. Clinical judgement, not statistics, made the call — the scales were judged heuristic rather than measurable.
Two papers appeared in Archives of General Psychiatry in November 1989. The first, on development, use and reliability, reported a study of four raters and forty patients at various stages of treatment: agreement was excellent for the total score and for each of the ten items, with strong internal consistency. The second paper addressed validity, drawing on pre-treatment assessment of forty-two patients. Every item was frequently endorsed and every item recorded scores across the severity range, which is the quiet achievement. The scale worked in people who were mildly affected and in people who had almost stopped functioning.
The instrument's most valuable output is often a sentence rather than a number. A patient who scores 19 because of a two-hour nightly checking ritual is a different clinical problem from one who scores 19 because of a single intrusive image that ruins every meal. The checklist separates them. The total does not.
Photo by Markus Winkler on Unsplash
How a score becomes a gate
Sixteen is the number that opens doors. A total at or above that line, the bottom of the moderate band, is a standard entry criterion for OCD treatment trials. Below it, you are not sick enough to be studied.
The harder conventions concern change. A fall of 35 per cent or more in the total, alongside a clinician's global rating of much or very much improved, is the usual definition of response. A fall of 25 to 34 per cent counts as partial response. Consensus work published in World Psychiatry in 2016 pushed remission toward a total of 12 or below, though 16 is also used as a threshold in some services.
Run the arithmetic. A patient who starts at 40 and improves by 35 per cent lands at 26 — still inside the severe band. Response and recovery are not the same claim, and the two get blurred in press releases with tiresome regularity.
This is where the scale's authority actually comes from. Regulators, sponsors and reviewers can argue about mechanism, side effects and cost once they have agreed on the number. Agreement about the number is what makes the argument possible at all.
The item that punished recovery
Item 4 asks how hard the patient fights the obsession. Item 9 asks the same about the compulsion. On the original scoring, a person who always resists scores zero; a person who has given up the fight scores four.
Exposure and response prevention, the psychological treatment with the strongest evidence base in OCD, teaches something close to the opposite of fighting. Patients are trained to let the intrusive thought be present without ritualising, without wrestling it to the ground. A successfully treated patient who reports less resistance can therefore be scored as worse. The item conflates effort with outcome, and it behaves oddly in exactly the patients who have done the work.
Critics noticed early. By the mid-1990s the resistance and control items were being described as the weakest in the scale, and later psychometric argument ran in favour of deleting them outright. There was a second, blunter problem: a ceiling. Forty points could not separate someone severe from someone whose life had effectively stopped.
The Yale-Brown Obsessive Compulsive Scale, Second Edition was Storch and colleagues' answer, published in 2010 after testing 130 treatment-seeking adults. Items were re-spaced onto a 0 to 5 range and the total moved to 0 to 50. Active avoidance was folded into severity rather than sitting outside it. The resistance-to-obsessions item was replaced with an obsession-free interval; resistance to compulsions stayed. The checklist was reworded and expanded. Reliability came out strong — internal consistency of .91 for the checklist and .89 for the severity scale, with interrater and test-retest correlations above .85 — though confirmatory analysis failed to fit any previously published factor structure.
One practical warning travels with the revision. A Y-BOCS-II total cannot be read against the 0 to 40 bands, and switching instruments mid-treatment produces apparent jumps that are artefacts of the form.
Children, dimensions, and the self-report question
Paediatric use needed its own version. The Children's Yale-Brown Obsessive Compulsive Scale was validated in 1997 on sixty-five children aged eight to seventeen. Internal consistency came in at .87, with agreement of .84 for the total and .91 for the obsession subscale. The compulsion subscale lagged at .68, and the authors noted that reliability shifts with the child's age and with the awkward business of merging what a parent reports with what a child reports.
The Dimensional Yale-Brown Obsessive Compulsive Scale, published in Molecular Psychiatry in 2006, took a different route. Rather than one severity figure, it scores six symptom dimensions — contamination and cleaning, harm and checking, symmetry and ordering, sexual and religious, hoarding, and a mixed category — in 137 patients aged six to sixty-nine across sites in the United States, Canada and Brazil. Its global rating correlated with the conventional Y-BOCS total at .82. Individual dimension scores turned out to be largely independent of one another, which is the useful part: severity in one domain says little about severity in another.
Self-report versions have existed since the early 1990s. In a small comparison of eighteen patients, paper, telephone and computerised administrations produced near-identical results, with correlations between .97 and .99. The self-report form is not simply the interview on paper. It lists fewer symptoms and drops the open-ended write-in options, which is where the unmentionable material usually surfaces.
Photo by Leo_Visions on Unsplash
What the number is now asked to do
In February 2009 the US Food and Drug Administration granted Medtronic a humanitarian device exemption for deep brain stimulation in chronic, severe, treatment-resistant OCD — the first such clearance for a psychiatric indication. The evidence behind it covered twenty-six patients at four centres with at least twelve months of stimulation, who averaged a 40 per cent reduction in symptoms at one year, with two-thirds showing marked functional improvement in psychological, social and occupational terms. The labelled indication is narrow: bilateral stimulation of the anterior limb of the internal capsule, in adults who have failed at least three serotonin reuptake inhibitors. European approval followed in July 2009, with eligibility framed as a Y-BOCS score above 30.
The exemption drew fire. A group led by Joseph Fins argued in Health Affairs in 2011 that the pathway put patients at risk and should be rescinded. Benjamin Greenberg, a principal investigator on the multi-site trial, offered the counterweight: in the real world there had been at most about forty OCD surgeries in the United States in a year, roughly fifteen of them implants, and recruiting ten patients for a trial took three years. A later surgical review found that access had actually fallen after approval, because insurers declined to pay. A score above 30 opens the door to surgery and then runs into a different gate entirely.
Psychedelic research has leaned on the same instrument. A 2025 challenge study in Comprehensive Psychiatry gave nineteen adults 1 mg of psilocybin and then 10 mg four weeks apart, in fixed order, with raters blinded to sequence. The 10 mg dose produced a between-dose effect on the Y-BOCS of d = 0.82 one week later, driven mainly by the compulsion subscale (d = 0.74); the obsession subscale did not reach significance. The effect faded over the following three weeks. There was no randomisation, and the sample was small.
A separate phase 2 trial of twenty-eight adults with treatment-resistant OCD, controlled against niacin and reported with an acute version of the scale, found scores falling from about 24 to about 14 within forty-eight hours, with roughly 69 per cent meeting the 35 per cent response threshold at one week against none in the control arm. Note what had to happen for that result to be visible. The standard instrument looks back over the past week. A drug that works in two days forced the creation of an acute variant, because the scale's recall window was slower than the pharmacology.
The promise a scale can keep
Measurement in psychiatry has a habit of promising objectivity and delivering coordination instead. That is not a small thing. A trial that cannot coordinate cannot recruit, and a clinic that cannot coordinate cannot compare its waiting list against its outcomes.
The Y-BOCS has held its position because it does the coordinating job without pretending to do the metaphysical one. It does not tell you what an obsession is. It tells you that the person in Leeds and the person in Seoul are being asked the same question in the same order, and that the answers can be added up.
A scale is a promise that two strangers will count the same way. The Yale-Brown has kept that promise for nearly four decades. It never promised that the number and the life behind it would be the same size.
- Yale-Brown Obsessive Compulsive Scale
- Y-BOCS
- OCD research
- clinical assessment
- Y-BOCS-II

