Dyslexia is a specific impairment in learning to read and has strong heritability. An intronic deletion within the DCDC2 gene, with ~8% frequency in European populations, is increasingly used as a marker for dyslexia in neuroimaging and behavioural studies. At a mechanistic level, this deletion has been proposed to influence sensory processing capacity, and in particular sensitivity to visual coherent motion. Our re-assessment of the literature, however, did not reveal strong support for a role of this specific deletion in dyslexia. We also analysed data from five distinct cohorts, enriched for individuals with dyslexia, and did not identify any signal indicative of associations for the DCDC2 deletion with reading-related measures including in a combined samples (N =526). We conducted the first replication analysis for a proposed deletion effect on visual motion perception and found no association (N = 445 siblings). We also report that the DCDC2 deletion has a frequency of 37.6% in a cohort representative of the general population recruited in Hong Kong (N = 220). This figure, together with a lack of association between the deletion and reading abilities in this cohort, indicates the low likelihood of a direct deletion effect on reading skills. Therefore, on the basis of multiple strands of evidence, we conclude that the DCDC2 deletion is not a strong risk factor for dyslexia. Our analyses and literature re-evaluation are important for interpreting current developments within multidisciplinary studies of dyslexia and, more generally, contribute to current discussions about the importance of reproducibility in science.