The priming effects of graphemically similar (e.g., HOSE) and graphemi
cally dissimilar (e.g., Rows) rhymes on the naming of target words (NO
SE) were examined at prime-target stimulus onset asynchronies (SOA) of
36, 70, and 250 msec A four-field presentation procedure was used of
mask-prime-mask-target. The effects of rhyme primes were measured rela
tive to those of nonrhyming control primes (CHEF, DISH) that matched t
he rhymes in frequency and length and shared no letters in the same po
sitions. At SOAs of 36 and 70 msec, rhyme priming was inhibitory and e
qual for graphemically similar and graphemically dissimilar rhymes. At
SOA = 250 msec, rhyme priming was insignificant with a tendency towar
d facilitation. The results are discussed in the context of (1) contra
sting effects of complete versus partial phonological overlap within a
prime-target pair, and (2) the hypothesis that phonological codes sta
bilize fastest and provide, therefore, the earliest and major constrai
nt on word recognition.