This article discusses the limitations of using the I² statistic to assess heterogeneity in meta-analysis. The authors argue that I², which represents the percentage of variability in treatment estimates due to heterogeneity rather than sampling error, can be misleading when study sizes increase. As study sizes grow, I² tends to increase rapidly, approaching 100%, even if the true heterogeneity remains constant. This is because I² depends on study precision, which is inversely related to study size. Therefore, I² may not be an appropriate measure for assessing clinically relevant heterogeneity. The article presents a simulation study that demonstrates how I² increases with study size, even when the underlying heterogeneity remains the same. It also shows that in a sample of 157 meta-analyses, I² is strongly associated with study size. The authors emphasize that τ², the between-study variance, is a more appropriate measure for assessing heterogeneity in a clinical context, as it reflects the true variability between studies rather than sampling error. The authors also highlight that the interpretation of I² as a measure of clinical heterogeneity is flawed. They argue that I² should not be used as the sole criterion for deciding whether to pool studies in a meta-analysis. Instead, the clinical relevance of any heterogeneity should be considered. The article concludes that τ² is the appropriate measure for this purpose, as it provides a more accurate reflection of the true variability between studies.This article discusses the limitations of using the I² statistic to assess heterogeneity in meta-analysis. The authors argue that I², which represents the percentage of variability in treatment estimates due to heterogeneity rather than sampling error, can be misleading when study sizes increase. As study sizes grow, I² tends to increase rapidly, approaching 100%, even if the true heterogeneity remains constant. This is because I² depends on study precision, which is inversely related to study size. Therefore, I² may not be an appropriate measure for assessing clinically relevant heterogeneity. The article presents a simulation study that demonstrates how I² increases with study size, even when the underlying heterogeneity remains the same. It also shows that in a sample of 157 meta-analyses, I² is strongly associated with study size. The authors emphasize that τ², the between-study variance, is a more appropriate measure for assessing heterogeneity in a clinical context, as it reflects the true variability between studies rather than sampling error. The authors also highlight that the interpretation of I² as a measure of clinical heterogeneity is flawed. They argue that I² should not be used as the sole criterion for deciding whether to pool studies in a meta-analysis. Instead, the clinical relevance of any heterogeneity should be considered. The article concludes that τ² is the appropriate measure for this purpose, as it provides a more accurate reflection of the true variability between studies.