In a large epidemiological sample of young adults, predictions of the right-shift (RS) theory of Annett that cognitive abilities will vary with right-left hand skill were tested. Presenting a theory of a genetic balanced polymorphism with heterozygote advantage for laterality and ability, Annett & Manning (1989, 1990a) and Annett (1993c) claimed that probands at the right end of the R-L hand skill continuum would show lower general intelligence in IQ testing and that specific verbal abilities and educational success would be lower at both extremes of the R-L distribution, taking the form of an inverted U. Most of these predictions could not be confirmed by our study. In particular, our data contrast with the important and specific prediction of the RS theory that strong dextrals will be the most disfavoured group. In our sample, probands at the left end of the R-L continuum had significantly lower scores in spelling and educational success and showed a tendency to have lower non-verbal IQ scores, while strong dextrals tended to have average or even marginally higher ability scores. The effects, however, are small and decrease when controlling for other variables. Implications of these empirical findings for the right-shift theory are discussed.