Observers named visually presented words as quickly as possible when the words were presented to the centre of a viewing screen (Experiments 1, 5, and 6) or were flashed to either the left visual field/right hemisphere (LVF/RH) or right visual field/left hemisphere (RVF/LH) on each trial (Experiments, 2, 3, and 4). Words varied in frequency of occurrence in the language and in regularity of pronunciation. On lateralised trials, there was a RVF/LH advantage for both error rate and reaction time, with this hemispheric asymmetry for word naming being independent of both frequency and regularity. The RVF/LH advantage was also independent of whether the letters within a word were arranged horizontally or vertically. Error rates and reaction times were lower for high-than for low-frequency words and lower for words with a regular pronunciation than for exception words. In addition, on centre trials the effects of regularity were larger for low-frequency words than for high-frequency words. However, when words were presented in the visual periphery, the effects of regularity were of the same magnitude for high-and low-frequency words. In view of the theoretical importance of the Frequency by Regularity interaction, this latter result suggests that word processing is qualitatively different in the visual periphery than in the centre of vision; indicating that perceptual asymmetries in a typical visual half-field experiment may be limited in what they can tell us about the relative contributions of the two hemispheres to processing words in the centre of vision.