BackgroundA consensus on normal atrial deformation measurements by feature‐tracking cardiac MRI remained absent.PurposeProvide reference ranges for atrial strain parameters in normal subjects, evaluating the influence of field strength and analysis software on the measurements.Study TypeMeta‐analysis.Population2708 subjects from 42 studies undergoing cardiac MRI.AssessmentA systematic search was conducted from database (PubMed, Web of Science, ScienceDirect, and EMBASE) inception through August 2023. The random‐effects model was used to pool the means of biatrial strain parameters. Heterogeneity and clinical variable effects were assessed. Strain measurements among different field strengths and analysis software were compared.Statistical TestsThe inverse‐variance method, Cochrane Q statistic, and I2 value, meta‐regression analysis, and ANOVA were used; P < 0.05 was considered statistically significant.ResultsThe pooled means of left atrial (LA) total strain (εs), passive strain (εe), and active strain (εa) were 37.46%, 22.73%, and 16.24%, respectively, and the pooled means of LA total strain rate (SRs), passive strain rate (SRe), and active strain rate (SRa) were 1.66, −1.95, and −1.83, indicating significant heterogeneity. The pooled means of right atrial (RA) εs, εe, and εa were 44.87%, 26.05%, and 18.83%. RA SRs, SRe, and SRa were 1.66, −1.95, and −1.83, respectively. The meta‐regression identified age as significantly associated with LA εs, εe and SRe, field strength was associated with LA SRa (all P < 0.05). ANOVA revealed differences in LA εa and SRa among different analysis software and in LA εs and all LA strain rates (all P < 0.05) among field strengths. No significant differences were identified in RA strain across analysis software (RA strain: P = 0.145–0.749; RA strain rates: P = 0.073–0.744) and field strengths (RA strain: P = 0.641–0.794; RA strain rates: P = 0.204–0.458).Data ConclusionThis study demonstrated the pooled reference values of biatrial strain. Age, analysis software, and field strength were attributed to differences in LA strain, whereas RA strain showed consistency across different field strengths and analysis software. Limited study subjects may account for the absence of influence on RA strain.Level of Evidence1.Technical EfficacyStage 5.