To demonstrate an interfering effect of subvocal articulation on otherwise silent reading, college-student subjects were asked to repeatedly read, either silently or aloud, both tongue-twister sentences and control sentences matched for syntactic complexity, syllable count, and sentential stress pattern. A technique was developed to measure the amount of time needed for each repetition of a sentence whether done silently or aloud. A significant difference in reading time for tongue twisters as compared to their matched controls was found for both silent as well as out-loud reading. A variety of different kinds of articulatory errors occurred in the oral repetitions, and the number of such errors was highly correlated with oral reading time. While errors could not be measured in silent reading, this correlation suggested that comparable articulatory disturbances accounted for the slower time to silently repeat tongue twisters.