Armenian karmir “red” has often been considered as deriving from East Iranian, thus speaking in favour of relations between Armenian and Sogdian, a Middle Iranian language spoken at considerable distance from Armenia. For the origin of Hebrew karmīl, on the other hand, a Middle Persian “karmīr” has been suggested. In either case, the etymology would be Proto-Indo-European *ku̯ṛ́mi- “worm” (be it directly or as a borrowing from Sanskrit kṛ́mi-) from which the colour term would be derived in a way parallel to French vermeil “scarlet” from ver “worm”, thus a term referring to a red dye obtained from scale insects (cochineals). I argue that karmīr is not a Middle Persian word for “red”, that Sogdian is unlikely to be the source of the Armenian and Hebrew words, and that an Indian origin is not probable either because of the specific features of the Indian scale insect dye. Conversely, Armenian scarlet was widely known and appreciated already in antiquity, so that, for historical as well as linguistic reasons, the origin of the word is likely to be an Iranian language within the region where Armenian was spoken.