Like the Quenya syncope, a short vowel of the same quality as the preceding syllable was lost in Ilkorin, for example: ᴹ✶kwenedē > Dor. cwend (EtyAC/KWEN(ED)) or ᴹ✶thóron[ē] > Ilk. thorn (Ety/THOR). There is one example indicating that this change could occur even when the second primitive vowel was stressed: ᴹ✶salák-(wē) > Ilk. salch (Ety/SALÁK). In this case, though, it is likely that the stress shifted to the initial syllable before the vowel was lost.

There are a couple of counter-examples to this rule. First, the syncope did not occur when the result would be a cluster of three consonants: ᴹ✶Utubnu > Ilk. Uduvon (Ety/TUB), although it seems that if the third consonant was the semi-vowel [w], the syncope could still occur (see salch above). Second, it seems that the syncope did not occur when the second vowel was in the final syllable: ᴹ✶bóron- > Ilk. boron (Ety/BOR); this exception may apply to Ilk. aman as well (Ety/AM¹). In these last two examples, the syncope did apply to the plurals: burnin and emnin, where the lost vowel would no longer be in the last syllable.

It seems likely that the syncope occurred before voiced stops spirantalized after vowels, since ᴹ✶gálad[ā] > Ilk. gald (Ety/GALAD) and ᴹ✶ñgólod[ō] > Ilk. (n)gold (Ety/ÑGOLOD) rather than **galdh [galð] or **goldh [golð]. However, it is possible the syncope occurred afterwards, and voiced spirants were re-stopped in these positions.