Radloff’s CES-D in one page: 20 items on depressive symptoms, scoring, why 16 is an arbitrary cut-off, the Spanish cut-off of 26 and how to cite it.
Twenty items on how often depressive symptoms occurred over the past week, designed for epidemiological studies in the general population rather than for diagnosis. Freely available, with a Spanish adaptation validated against clinical interview.
By Víctor Ciudad-Fernández, PhD · Updated: October 6, 2026 · Leer esta ficha en español · Instruments
Lenore Radloff developed the CES-D at the Center for Epidemiologic Studies of the US National Institute of Mental Health (NIMH) to measure depressive symptoms in general-population studies, unlike the depression scales of the time, which were used mainly for diagnosis at intake or to track severity during treatment. It targets the current level of symptoms, with emphasis on the affective component, depressed mood.
Items were selected from previously validated, longer scales to cover the major components of depression identified in the clinical literature and factor-analytic studies: depressed mood, feelings of guilt and worthlessness, helplessness and hopelessness, psychomotor retardation, loss of appetite and sleep disturbance. Four items were worded positively to break response sets and, at the same time, capture positive affect or its absence.
Radloff was explicit about its limits: it is not a clinical diagnostic tool, individual scores should not be interpreted, and even group means should be read as levels of symptoms, not rates of illness. That is why, in research, the CES-D works better as a continuous measure than as a case counter.
Each item asks how often you felt that way during the past week and is scored from 0 to 3 by frequency. The four positive items are reverse-scored before summing, and the total ranges from 0 to 60: higher scores mean more symptoms, weighted by frequency.
In the original study it was given as part of a household interview and also self-administered by mail. Radloff flagged possible interviewer and format effects on mean scores, and recommended caution with bilingual respondents and with anyone who might misunderstand the items.
In the general-population samples of the original study, principal components analysis found four factors that together explained 48% of the variance, with a very stable loading pattern across samples: depressed affect, positive affect, somatic and retarded activity, and interpersonal. Even so, the scale is usually scored as a single total.
The four-factor structure has been debated: Carleton et al. (2013) review the proposed models and test their fit in five samples. In the Spanish adaptation, exploratory analysis also found four factors, explaining 55.9% of the variance.
The cut-off of 16 does appear in the original paper, but as one Radloff herself calls arbitrary, used to describe how many people exceeded it, and in the same paper she warns that appropriate cut-offs for clinical screening were yet to be validated.
| Score | Use | Source |
|---|---|---|
| ≥16 | Arbitrary cut-off in the original study: exceeded by 70% of patients and 21% of the population | Radloff (1977) |
| ≥26 | Screening in the Spanish non-psychiatric population aged 18 to 34 (sensitivity .906, specificity .918) | Vázquez, Blanco & López (2007) |
The difference between 16 and 26 is not a detail: at 16, one in five people in the general population screens “positive”. Choose the cut-off for your population and purpose, and justify it with its reference.
A screening questionnaire is not a diagnosis: a high score means a clinical assessment is warranted, not that a disorder is present.
Vázquez, Blanco and López (2007) adapted the CES-D for the Spanish non-psychiatric population: they gave an interviewer-administered translation to 554 people aged 18 to 34 and compared their scores with diagnoses by expert clinicians based on the SCID-CV. Internal consistency was satisfactory (α = .89), exploratory analysis yielded four factors and a cut-off of 26 proved suitable for screening, with sensitivity and specificity above .90.
That cut-off is much higher than the commonly quoted 16, which underlines Radloff’s warning: a screening cut-off has to be validated in each population.
License: Freely available: it was developed at the NIMH Center for Epidemiologic Studies, and the literature describes it as a freely available measure (Carleton et al., 2013). A publication must cite the original paper and the version you use.
The 20 items are in Radloff’s (1977) original paper, which the University of Minnesota repository allows students and faculty to reproduce at no cost for academic use. The Spanish adaptation is in Psychiatry Research.
Cite the original paper and, if you use the Spanish version, its adaptation:
@article{radloff1977cesd,
author = {Radloff, Lenore Sawyer},
title = {The {CES-D} {Scale}: A self-report depression scale for research in the general population},
journal = {Applied Psychological Measurement},
year = {1977},
volume = {1},
number = {3},
pages = {385--401},
doi = {10.1177/014662167700100306}
}The four positive items cause most of the trouble: if one was not reversed, or if in your sample they behave as a separate factor, it shows up in their correlation with the rest. The item analysis calculator gives you each item’s discrimination and alpha-if-item-deleted before you sum.
Sixteen is the arbitrary cut-off from the original paper; with it, about one in five people in the general population crosses the threshold. In the Spanish non-psychiatric population aged 18 to 34, the cut-off validated against clinical interview was 26. If your sample resembles that one, use 26 and cite Vázquez et al. (2007); otherwise, consider working with the continuous score.
It measures depressive symptoms over the past week. Radloff warned that it is not a diagnostic tool and that scores should be read as levels of symptoms, not as the presence of a disorder.
This fact sheet is an academic reference for information purposes: it does not reproduce any questionnaire item. Using the instrument in your research or practice may require citing its authors or asking for permission: see “How to get the instrument”.