OCI-R: 18 items covering six kinds of obsessive-compulsive symptoms
The Obsessive-Compulsive Inventory originally had 84 items. Foa and colleagues cut it to 18 in 2002, and the short version became the one most studies use. It asks about the past month and rates how much each experience bothered you.
Answers run 0 (not at all) to 4 (extremely). Nothing is reverse-scored. A subscale score runs 0 to 12 and the total runs 0 to 72. It takes about 5 to 10 minutes and is written for adults aged 18 and over.
Six subscales of three items
| Subscale | Items | What it covers |
|---|---|---|
| Washing | 5, 11, 17 | Contamination fears and excessive washing |
| Checking | 2, 8, 14 | Repeated checking of locks, taps and switches |
| Ordering | 3, 9, 15 | Needing symmetry and precise arrangement |
| Obsessing | 6, 12, 18 | Intrusive thoughts that will not let go |
| Hoarding | 1, 7, 13 | Collecting things and difficulty discarding |
| Neutralizing | 4, 10, 16 | Counting or mental acts used to reduce anxiety |
When DSM-5 moved hoarding disorder out of OCD and made it a separate diagnosis, the scale adapted. The 15 items covering the other five subscales can be scored as OCI-OCD, and the three hoarding items as OCI-HD. A 2015 psychometric evaluation confirmed both scores hold up, and ROC analysis showed the cutoffs keep usable sensitivity and specificity.
Cutoffs
| Score | Threshold |
|---|---|
| Total | 18 or higher suggests possible OCD |
| Any subscale mean | 2.5 or higher suggests possible OCD |
| Total, contemporary sources | 21 or higher suggests probable clinical OCD |
Those lines come from the original manual and from later work, and they are screening thresholds rather than diagnostic ones. A 2020 study produced severity benchmarks and clinical norms for reading a total as mild through severe instead of as a yes or no.
Reliability
Internal consistency for the total runs .86 to .95 across samples. The three-item subscales land between .70 and .90 depending on the subscale and the sample, with neutralizing the weakest in some of them. Retest correlations over intervals from one week to three months fall between .70 and .85.
What it measures well, and what it misses
Factor analysis supports six factors, or a five-plus-one arrangement with hoarding separate. Scores correlate moderately to strongly with the Y-BOCS and other OCD measures, and the scale separates people with OCD from people with other anxiety disorders and from controls.
It is also sensitive to treatment, which makes it useful for tracking change during exposure and response prevention. The limits are the usual ones for self-report: underreporting when symptoms feel shameful, and social desirability pulling scores down. A questionnaire cannot tell the difference between a compulsion someone has and one they are describing from memory.
Practical notes
The scale has been translated into 13 languages and is copyrighted by Edna B. Foa. The item wording is published in the manual rather than released into the public domain.
Scoring high on one subscale does not make a diagnosis, and scoring low does not rule one out. Obsessive-compulsive symptoms exist on a continuum in the general population, and plenty of people have some checking or ordering habits without having OCD. If symptoms cause distress, or interfere with work or relationships, that is worth raising with a clinician. This page is not a diagnosis and does not replace one.