While there is growing evidence that some dyslexic children suffer from a deficit in simultaneously processing multiple visually displayed elements, the precise nature of the deficit remains largely unclear. The aim of the present study is to investigate possible cognitive impairments at the source of this deficit in dyslexic children. The visual processing of simultaneously presented letters was thus thoroughly assessed in two dyslexic children by means of a task that requires the report of briefly presented multi-letters arrays. A computational model of the attentional involvement in multi-object recognition (Bundesen, 1990, 1998) served as framework for analysing the data. By combining psychophysical measurements with computational modelling, we demonstrated that the visual processing deficit of simultaneously displayed letters, observed in the two dyslexic individuals reported in the current study, stems from at least two distinct cognitive sources: a reduction of the rate of-letter-information uptake, and a limitation of the maximal number of elements extracted from a brief visual display and stored in visual short-term memory. Possible relations between these impairments and learning to read proficiently are discussed.