SRMR (Standardized Root Mean Square Residual)
A good SRMR is .08 or lower. SRMR is the standardized average difference between the observed and model-implied correlations.
srmrWhat it means
SRMR summarizes how far the correlations your model predicts are from the correlations actually in the data. A value of 0 means a perfect match; larger values mean bigger residuals. It is intuitive because it is on the scale of correlations.
Why it matters
SRMR is a useful complement to RMSEA and CFI because it focuses directly on residuals, where misspecification shows up. It is part of the standard fit set for CFA and SEM.
How to report it
Report SRMR to three decimals. The common guideline is SRMR ≤ .08 for good fit (Hu & Bentler, 1999; https://doi.org/10.1080/10705519909540118), and up to .10 is often treated as acceptable (Schermelleh-Engel, Moosbrugger & Müller, 2003). Check it with the fit index checker.
Reference table
| Cutoff | Interpretation |
|---|---|
| ≤ .08 | Acceptable fit |
| 0 | Perfect match |
Frequently asked questions
What is a good SRMR value?
SRMR at or below .08 is the usual guideline for good fit (Hu & Bentler, 1999), and up to .10 is often treated as acceptable (Schermelleh-Engel et al., 2003). Lower values indicate the model reproduces the observed correlations more closely.
What does a high SRMR mean?
A high SRMR means the model-implied correlations differ noticeably from the observed ones, which signals misspecification. Inspect the residual correlation matrix to see which relationships are poorly reproduced.
How to cite this page
If this page helped your research, you can cite it (APA 7):
Szymkowiak, A. (2026). SRMR (Standardized Root Mean Square Residual). Explonia. https://explonia.com/glossary/srmr/
Last updated: October 2026
Check these values on your own data
Upload your data file (CSV or Excel) and get the same statistics, with a draft APA paragraph that you edit. Free to start.
Analyze your data