This study tested a sample of 63 children twice in a longitudinal design over 14 months to examine their theory-of-mind (ToM) understanding in relation to their number of child-aged siblings (1-12 years). Age-appropriate batteries of ToM tests emphasising false belief were given at the start of the study, when children had a mean age of 4-2 (range: 3-3 to 5-6), and at the end, when mean age was 5-4 (range: 4-3 to 6-9). Irrespective of chronological age, children with 2 or more child siblings scored significantly higher on both the earlier and the later battery than those with no child-aged siblings. Hierarchical multiple regression analyses revealed that, over and above chronological age and verbal intelligence, having more child siblings predicted higher ToM scores at Times 1 and 2. Furthermore, at Time 2, the participant's number of child-aged siblings continued to predict higher ToM scores even after controlling for age, verbal intelligence, and Time I ToM scores. Results were considered in relation to the kinds of family-based social and conversational experiences that might foster ToM growth throughout the period from toddlerhood to the threshold of primary school. (C) 2006 Elsevier Inc. All rights reserved.