This dataset comprises 40 digitised Hebrew manuscripts (900 – 1747; unknown date), with their shelfmarks in alphabetical order (Or 2510 – Or 2588). These manuscripts are out of copyright. These JPEG files were converted from TIFF files using IrfanView, and then further compressed using JPEGMini. They can be used for image and text analysis, teaching, art projects, or anything else of inspiration. We would appreciate it if users could read our Ethical terms of use guide before reusing our Hebrew manuscripts datasets: https://www.bl.uk/help/ethical-terms-of-use