Nursery teachers assessed 234 children aged 4 1 2 to 6 1 2 years on the DSM-IV criteria for hyperactivity-impulsivity and attention deficit (ADHD). Using slightly lenient classification criteria 21 children scored on hyperactivity-impulsivity and/or attention deficit and were classified as 'at risk of ADHD'. An age-matched control group showed significant relations between acquisition of an advanced theory of mind, in particular understanding second-order beliefs, and executive competence as measured by several tasks from the NEPSY. This extends findings from younger children that theory of mind development relates to executive control . The group at risk of ADHD showed impaired performance relative to the control group on several executive tasks but no impairment at all on the advanced theory of mind tasks. This speaks against the theory that later theory of mind development is a consequence of improvements in executive control. Copyright # 2002 John Wiley & Sons, Ltd.Key words: hyperactivity-impulsivity; attention deficit (ADHD); theory of mind development; executive control This study pursued three objectives. The first objective is to follow up and extend an earlier study by Hughes et al. (1998) assessing 'hard to manage' 4-year-old preschoolers' impairment on executive and theory of mind tasks. Only three studies have so far explored the association between theory of mind abilities and executive functions in 6-to 10-year-old ADHD children (Charman et al., in press) or 'hard to manage' pre-schoolers (Hughes et al., 1998;Speltz et al., 1999). The study by Hughes et al. (1998) compared 'hard to manage (H2 M)' pre-schoolers with controls on a large test battery assessing theory of mind, emotion understanding and executive function. There was no difference between groups on false-belief prediction and explanation tasks once verbal ability had been controlled. There was, however, poorer affective perspective-taking and inhibitory control in the H2 M group than in the control group even after verbal ability and social background had been taken into account. Moreover, significant correlations using aggregate scores were found between theory of mind or emotion understanding and executive function in the H2 M group but not in the