The study evaluated a substantially updated version of Orton's (1937) "classical" idea of a significant relationship in dyslexic children between cerebral lateralization and their word decoding deficits. Attentional lateralization was examined under the assumption that covert spatial attention when directed contralaterally interacts with age-invariant cerebral asymmetries for receptive speech. Thirty dysphonetic dyslexic children were compared to 30 younger normal readers who were matched to the dyslexics in reading comprehension. The children were tested in left ear (LE) and right ear (RE) directed attention dichotic listening (DAD), and in pseudoword decoding, word recognition, reading comprehension, spelling, arithmetic, and in general intelligence (IQ). Group comparisons in DAD failed to show any differences, confirming the mounting evidence that dyslexia is not related to incomplete lateralization. Entering the DAD scores of the dyslexics (LE first, LE second, RE first, RE second) as predictors of achievement revealed that, independently of chronological age (CA) and IQ, their ability to recall items from the LE first produced a negative regression which predicted 42 percent of the variance in pseudoword decoding. Selective report from the LE also produced small but significant negative correlations with visual recognition of real words and spelling; but no relationship to reading comprehension. IQ was related to reading comprehension and to the ability to shift attention from the LE to the RE. Even though the dyslexics were lateralized normally, weak lateralization was related specifically to phonological word decoding, a core deficit in dyslexia. However, unlike Orton's concept, these findings suggest that dyslexics suffer from exuberant right hemisphere processing in response to spatial attentional demands that, in turn, interferes transcallosally with the development of the sound-symbol representations that are required for fluent reading. Lateralization, per se, is unaffected by the disorder.