Twenty male and twenty female fourth graders were randomly selected from a group of black English speakers, and another twenty male and twenty female fourth graders were randomly selected from a group of standard English speakers. Dialect groups were distinguished by use of the Baratz Sentence Repetition method (Baratz, 1969). Subjects were tested for sensitivity to phonological information by the use of the Conrad paradigm (Mark, Shankweiler, Liberman, & Fowler, 1977) and for levels of reading automaticity by use of the Corcoran paradigm (Drewnowski, 1978). Speakers of black English were found to display low sensitivity to phonological information and less automaticity at the word level and the level of connected discourse than speakers of standard English. No differences between the dialect groups was found in reading automaticity at the letter or letter group level. Analysis of covariance was used to control for differences between the dialect groups in reading proficiency. A significant proportion of the variance in reading proficiency was accounted for by automaticity at the level of connected discourse. No differences were found as a function of the ages or sexes of the subjects. The author concluded that black English dialect is associated with low sensitivity to phonological information with consequent low automaticity in word-level reading skills. He argued that there was good reason to reconsider the linguistic interference hypothesis in revised form