Hi, and thanks for all of the input - I think I'm beginning to grok it. > On second thought you could probably use NFD normalization to separate > base letters from accents, uppercase the base letters and then > (optionally) NFC normalize everything again. Still insane ;-). As far as I can see, I can only do what I want by using the following. If there's a monthly prize on this list for the most insanely contrived piece of code, I think this is a strong contender: test.x = 'abc' SELECT UPPER(REGEXP_REPLACE(x, '(.)(.)(.)', '\1')) || UPPER(REGEXP_REPLACE(x, '(.)(.)(.)', '\2')) || UPPER(REGEXP_REPLACE(x, '(.)(.)(.)', '\3')) FROM test; Result: ABC (see https://dbfiddle.uk/LQ-6udga). Still a bit of work to do, but I'm getting there :-) E. > hp
