The Revised Child Anxiety and Depression Scale – Child version (RCADS-Child) is a 47-item self-report questionnaire designed to assess symptoms of anxiety and depression in children and adolescents aged 8 to 18 years.
The RCADS-Child (Chorpita et al., 2000) consists of six subscales that screen for high prevalence disorders in youth:
In addition to the subscales, a Total Internalising score that is a combination of all six subscales is provided that gives an overall indication of the level of internalising problems. A Total Anxiety score that is a combination of the five anxiety subscales provides an overall indication of anxiety problems.
A parallel parent-report version, the RCADS-Parent, can be used in conjunction with the RCADS-Child to gather additional clinical information. The RCADS is well-suited for screening and monitoring of anxiety and depressive symptoms in youth.
Example RCADS-Child Items:
The RCADS-Child yields a Total Internalising Scale (sum of all 6 subscales) and a Total Anxiety Scale (sum of 5 anxiety subscales) to provide an overall indication of symptom severity. The RCADS provides raw scores for the six subscales that are converted to percentiles based on the child’s age group (8-12 or 13-18 years) from the Australian normative data. A percentile of 50 reflects the average symptom level relative to same-aged peers, while increasingly higher percentiles indicate more severe symptoms (e.g. > 93.31st percentile signifies elevated symptoms).
A descriptor is provided based upon the following criteria:
The six subscales assessed by the RCADS-Child are:
The RCADS subscales are best conceived as dimensional symptom clusters rather than diagnostic categories. Higher scores suggest greater symptom severity and clinical risk and should be integrated with clinical judgement to guide diagnostic formulation and treatment planning.
Plots demonstrating the subscale percentiles are presented upon first administration. Longitudinal plots are displayed for both the subscale percentiles and the Total Internalising and Anxiety percentiles upon multiple administrations to enable tracking of symptoms over time.
The RCADS was developed using anxiety items from the Spence Children’s Anxiety Scale (Spence, 1997) and adding a depression scale, then validated in a large community sample of children and adolescents (N = 1,641) by Chorpita et al. (2000). Exploratory factor analysis yielded a 6-factor solution consistent with DSM-IV anxiety and depressive disorders. The measure has demonstrated good internal consistency for all subscales (α = .71-.85) and the Total Internalising Scale (α = .93-.96). Test-retest reliability was also satisfactory over a 1-week interval (rs = .65-.80; Chorpita et al., 2000).
Confirmatory factor analysis in separate child and adolescent samples supported the 6-factor structure, with acceptable goodness-of-fit (Chorpita et al., 2005). The RCADS has shown good convergent and discriminant validity with other established measures of child anxiety and depression (Chorpita et al., 2005; Ebesutani et al., 2010, 2012).
The psychometric properties of the RCADS have been further examined in an Australian community sample of youth aged 8-18 (N = 405; de Ross et al., 2002). Results mirrored those of Chorpita et al. (2000), with the 6-factor model exhibiting satisfactory fit. The Australian sample was split into younger (8-12) and older (13-18) age groups to provide normative and percentile data for contextualising scores (if used with clients outside of this age range then percentiles are calculated using combined 8-18 years data).
The percentiles for the Total Internalising Score and the Total Anxiety Score were interpolated using the de Ross et al. (2002) data for subscale scores. This was done using percentile intervals of 5 (i.e., 5, 10, 15…) where the score for each subscale was calculated using subscale means and standard deviations and summed to provide an estimate of the combined score for each percentile interval. Percentiles between the intervals were then interpolated using a linear approach to determine corresponding percentiles for each possible score.
Descriptors are based upon T-Scores, however for ease of interpretation these scores are converted to percentiles:
Note the ‘Mild’ descriptor range was added by NovoPsych to be consistent with the interpretation of the Spence Children’s Anxiety Scale (Spence, 1997), from which the RCADS-Child was derived, where scores above the 84th percentile are classified as clinically significant anxiety symptoms.
A reliable change score has been determined by NovoPsych for the Internalising Score from data collected between July 2020 and June 2024 where clients had completed multiple RCADS-Child assessments. All data was included in the analysis, so no data was removed based upon initial symptom severity. The resultant sample size was 7,695. The reliable change score was calculated using the Jacobson-Truax method of clinical significance classification (Jacobson & Truax, 1991) using calculated Cronbach alpha’s from the NovoPsych data. The internal reliability calculated for the Internalising Score was 0.96 and the reliable change score calculated was 15.
For comprehensive information about the RCADS, see the RCADS website
The percentiles for the Total Internalising Score and the Total Anxiety Score were interpolated using the de Ross et al. (2002) data for subscale scores. This was done using percentile intervals of 5 (i.e., 5, 10, 15…) where the score for each subscale was calculated using subscale means and standard deviations and summed to provide an estimate of the combined score for each percentile interval. Percentiles between the intervals were then interpolated using a linear approach to determine corresponding percentiles for each possible score.
The percentile tables below demonstrate the interpolated values for the Total Internalising and Total Anxiety scores for the different age groups. The descriptors (i.e., Mild, Elevated, and Clinical) are highlighted in colour in the Percentile column.
This is a common and understandable point of confusion. The discrepancy does not reflect an error on either platform. It reflects a fundamental difference in the normative samples being used. NovoPsych benchmarks scores against the de Ross et al. (2002) Australian community sample (N = 405), whereas the UCLA scoring tool typically uses the Chorpita et al. (2000) US validation sample (N = 1,641). These are different populations with different base rates of anxiety symptom endorsement, so a given raw score will produce a different T-score, and therefore a different descriptor, depending on which reference group is used. For clinicians practising in Australia or the UK, the de Ross et al. (2002) Australian norms are likely to be more demographically appropriate than the Chorpita et al. (2000) norms, which were drawn entirely from schools in O’ahu, Hawaii, with a highly specific ethnic composition not representative of the broader English-speaking world.
The Mild descriptor is an addition made by NovoPsych and does not appear in the original RCADS scoring framework. It covers the range between the 85th and 93rd percentile (equivalent to T-scores between 60.01 and 64.99), which falls between the standard Normal and Elevated categories. This range was added to align with the Spence Children’s Anxiety Scale (SCAS), from which the RCADS was derived, where scores above the 84th percentile are considered clinically noteworthy. Without this intermediate category, clinicians using the traditional RCADS thresholds would classify scores in this range as Normal, potentially missing children whose symptoms, while not reaching the traditional Elevated threshold, are still meaningfully elevated relative to their peers. The Mild descriptor flags these scores as warranting clinical attention, consistent with the spirit of the SCAS scoring approach.
NovoPsych converts T-scores to percentiles because percentiles provide a more clinically intuitive and directly communicable comparison. A T-score is a standardised score with a mean of 50 and standard deviation of 10, which requires additional interpretation before its clinical meaning is clear. A percentile, by contrast, allows you to make a direct statement about where a child stands relative to their peers — for example, “this child scores higher than 96% of adolescents their age on Social Phobia.” This is more useful in report writing and in communicating results to clients, families, and other professionals.
This is a feature of normative scoring, not an error. Percentiles reflect how a score compares to the reference population, not how high a score is relative to the maximum possible. For Separation Anxiety in particular, mean scores in the 13–18 normative group are notably low (mean = 2.32, SD = 2.34 in the de Ross et al. (2002) sample), because separation anxiety is developmentally uncommon in adolescents. This means that even a modest absolute score can place a young person well above the majority of their peers. For example, a raw score of 9 for a 13-year-old on the Separation Anxiety subscale is approximately 2.9 standard deviations above the normative mean, which correctly corresponds to a very high percentile. This is clinically meaningful: it reflects that this level of separation anxiety is genuinely unusual for an adolescent, even if the raw score does not appear particularly large relative to the maximum possible score.
The 13–18 age band used in NovoPsych’s RCADS-Child scoring reflects the normative structure of the de Ross et al. (2002) Australian standardisation study, which split the sample into two groups: children aged 8–12 and adolescents aged 13–18. This grouping was not arbitrary. The data show that statistically significant age differences in anxiety and depression subscale scores are concentrated at the transition from childhood to adolescence, rather than within the adolescent years themselves. For Separation Anxiety specifically, mean scores in the 13–18 group are markedly lower and less variable than in the 8–12 group (means of 2.45 vs 4.23; SDs of 2.32 vs 4.34), consistent with separation anxiety becoming generally attenuated across adolescence as a whole. The primary normative reference group transition therefore occurs at the 12/13 boundary, and treating the 13–18 band as a single normative group is well supported by the available data. As always, clinicians should apply professional judgement when interpreting results for any individual client, particularly on subscales where developmental context is relevant.
This is expected behaviour, not an error. The RCADS-Child and RCADS-Parent are normed against entirely different standardisation samples, and each version is completed by a different informant – the child themselves versus a parent reporting on the child’s behaviour. Because these two groups naturally differ in how they perceive and endorse anxiety and depression symptoms, the normative distributions for a given subscale are not identical across versions. The RCADS-Child uses Australian community norms from de Ross et al. (2002), while the RCADS-Parent uses norms from Ebesutani et al. (2011), a separate community sample of 967 parents of children and adolescents. A raw score of 16 on Social Phobia, for example, will be benchmarked against a different mean and standard deviation depending on which version was administered, and will therefore produce a different percentile and potentially a different descriptor.
The RCADS-Parent also applies more fine-grained normative stratification than the RCADS-Child, using grade-based age bands in combination with the child’s recorded sex. If sex is not specified or non-binary is selected, a combined male/female norm across ages 8–18 is applied instead. This means that the client’s age and recorded gender will influence the resulting percentile on the parent version in ways that may differ from the child version. When administering both versions as part of a multi-informant assessment, it is therefore expected and clinically informative that the two versions may yield different descriptors for the same subscale. Discrepancies between parent and child report are themselves meaningful data about the consistency of symptom presentation across informants and contexts.
Bruce F. Chorpita and Susan H. Spence.
Chorpita, B. F., Yim, L., Moffitt, C., Umemoto, L. A., & Francis, S. E. (2000). Assessment of symptoms of DSM-IV anxiety and depression in children: A revised child anxiety and depression scale. Behaviour research and therapy, 38(8), 835-855. https://doi.org/10.1016/s0005-7967(99)00130-8
Chorpita, B. F., Moffitt, C. E., & Gray, J. (2005). Psychometric properties of the Revised Child Anxiety and Depression Scale in a clinical sample. Behaviour Research and Therapy, 43(3), 309-322. https://doi.org/10.1016/j.brat.2004.02.004
de Ross, R. L., Gullone, E., & Chorpita, B. F. (2002). The Revised Child Anxiety and Depression Scale: A psychometric investigation with Australian youth. Behaviour Change, 19(2), 90-101. https://doi.org/10.1375/bech.19.2.90
Ebesutani, C., Bernstein, A., Nakamura, B. J., Chorpita, B. F., & Weisz, J. R. (2010). A psychometric analysis of the revised child anxiety and depression scale—parent version in a clinical sample. Journal of Abnormal Child Psychology, 38(2), 249-260. https://doi.org/10.1007/s10802-009-9363-8
Ebesutani, C., Reise, S. P., Chorpita, B. F., Ale, C., Regan, J., Young, J., … & Weisz, J. R. (2012). The Revised Child Anxiety and Depression Scale-Short Version: Scale reduction via exploratory bifactor modeling of the broad anxiety factor. Psychological Assessment, 24(4), 833-845. https://doi.org/10.1037/a0027283
Spence S. H. (1997). Structure of anxiety symptoms among children: a confirmatory factor-analytic study. Journal of abnormal psychology, 106(2), 280–297. https://doi.org/10.1037//0021-843x.106.2.280
