From JRA-6445: It wasn't clear to me or my colleague when we looked at the importer that the importer only really worked with IDs. It would be good if the documentation could make that clearer. It might also be worth doing one of two things: 1) add a feature that automatically figures out the id based on text lookup (it is unlikely that people will create duplicate names for these types of things) avoiding the need for mapping at all in most cases. When setting up issue types etc people will often choose the same names as used in the system(s) they are migrating from. 2) Don't allow the user to proceed if the type required is an int id and the value in CSV file is a string. Even an error like "Type mismatch - integer expected in Issue Type field" would have led me to guess that mapping was required. >> Both would make excellent usability improvments for the CSV importer in the future.