We examined the interactions between four genes associated with dyslexia (ROBO1, KIAA0319, DCDC2, DYX1C1) and education on the tendency to become an entrepreneur. We used a two-staged design consisting of a discovery sample of 692 individuals, and a replication sample of 797 participants from the TwinsUK cohort. Associations were identified between entrepreneurship and interactions of education and ROBO1 rs654867 and KIAA0319 rs6902039 with the stage 1 samples. However these were not independently replicated and the associations were no longer significant when the samples from the 2 stages were combined. A tagging SNP approach was used to investigate the effect of the interactions between education and 191 tagging SNPs from the candidate genes on entrepreneurship. While we found several significant interactions (DCDC2, KIAA0319 and ROBO1), none passed the stringent threshold for significance of a Bonferroni correction. Similar to the case with other behavioural genetics phenotypes, large sample sizes will be required to identify significant gene-environment interactions in entrepreneurship after making Bonferroni corrections.