Two tests of spatial ability, mental paper folding given to 459 14–15-year-olds representative of the secondary school population, and recall of the Rey figure given to 428 undergraduates, found significant relationships between ability and hand preference when preferences were classified in subgroups of left- and right-handers (Annett, 1970a). In 14–15-year-olds and in female undergraduates there was a W-shaped relation between hand preference and spatial ability. The highest scores were in the centre for right-handers with strong sinistral tendencies. Male undergraduates showed a linear trend, with spatial ability highest in strong left-handers and declining from left to right across levels of hand preference. When spatial test scores were considered in relation to differences between the hands in skill a decline of ability with increasing dextrality was found in right-handers and in left-handers. In terms of the right shift theory (Annett, 1972, 1985), the evidence suggests that there are costs for spatial ability associated with the presence of the rs+ gene, even in single dose (rs + - genotypes), and that those who lack the gene (rs − − genotypes) have advantages for spatial ability (Annett, 1991c).