The most widely accepted cognitive theory of dyslexia is that of a phonological deficit, manifested by poor phonological awareness, poor verbal short-term memory and slow automatized naming. We will recall the main pieces of evidence in favour of a phonological deficit in dyslexia, and of its causal relationship with reading skills. We will then consider the possibility that other, independent cognitive deficits in the visual or visual/attentional domain, might explain a certain proportion of cases with dyslexia. We will report the results of a large French study in which we have concurrently investigated phonological skills, visual attention span, and visual stress.