The primary reflex system emerges during fetal life and is inhibited during the first year after birth. Our aim was to examine the effects of persistence of this early neurological system on the attainment of core literacy skills in dyslexic and non-dyslexic poor readers. We assessed the prevalence of a persistent primary reflex in a cross-sectional, representative sample of children (n=739) aged 7-9 years old attending mainstream primary school in Northern Ireland using standardised educational tests, and a clinical diagnostic test for a primary reflex (the asymmetrical tonic neck reflex (ATNR)). Multiple regression analyses, involving all of the sample children, revealed that persistence of the ATNR was significantly predictive of attainments in reading (t=-8.34, p<.001), spelling (t=-8.00, p<.001), non-word reading (t=-16.15, p<.001), and verbal IQ (t=-4.71, p<.001). ANOVA tests revealed that there were no differences between the performance of dyslexic and non-dyslexic poor readers on any of the outcome measures (reading (F(1, 289)=0.51, p=.48), spelling (F(1, 289)=0.02, p=.90), non-word reading (F(1, 289)=0.76, p=.38), ATNR level (F(1, 289)=2.54, p=.11)). Further ANOVA tests revealed that males had significantly higher levels of persistent reflex than females (F(1, 737)=15.21, p<.001), and that children from socially disadvantaged backgrounds had significantly higher levels of reflex than children who were not socially disadvantaged (F(1, 737)=20.84, p<.001). The findings suggest that for many children in mainstream schooling, the attainment of core educational skills may be affected by the persistence of a brainstem mediated reflex system that should have been inhibited in the first year after birth. Furthermore, these findings suggest that dyslexia is not a distinct category of poor reading, and that it may be more valid to term all poor readers as dyslexic irrespective of IQ.