This Perspective highlights the need for transparent thresholding rather than only showing results beyond fixed P values, to avoid the loss of valuable information and to improve the reporting of neuroimaging (and other scientific) results.