Cluster-extent based thresholding is widely used in fMRI studies for multiple comparisons correction due to its high sensitivity. However, it provides low spatial specificity, making it difficult to infer the significance of specific locations within clusters. Liberal primary thresholds (e.g., p < .01) are common, often leading to large clusters spanning multiple anatomical regions, which complicates interpretation. A survey of 814 fMRI studies found that liberal thresholds are prevalent, with software defaults heavily influencing their use. Simulations showed that liberal thresholds increase false positives, reduce localization accuracy, and hinder interpretation. To mitigate these issues, the authors recommend using stricter thresholds (e.g., p < .001), voxel-wise corrections, and transparent reporting of spatial precision. Alternative methods like threshold-free cluster enhancement (TFCE) and hierarchical FDR control are also suggested. The paper emphasizes the importance of choosing appropriate thresholds to ensure accurate and interpretable fMRI results.Cluster-extent based thresholding is widely used in fMRI studies for multiple comparisons correction due to its high sensitivity. However, it provides low spatial specificity, making it difficult to infer the significance of specific locations within clusters. Liberal primary thresholds (e.g., p < .01) are common, often leading to large clusters spanning multiple anatomical regions, which complicates interpretation. A survey of 814 fMRI studies found that liberal thresholds are prevalent, with software defaults heavily influencing their use. Simulations showed that liberal thresholds increase false positives, reduce localization accuracy, and hinder interpretation. To mitigate these issues, the authors recommend using stricter thresholds (e.g., p < .001), voxel-wise corrections, and transparent reporting of spatial precision. Alternative methods like threshold-free cluster enhancement (TFCE) and hierarchical FDR control are also suggested. The paper emphasizes the importance of choosing appropriate thresholds to ensure accurate and interpretable fMRI results.