INTRODUCTION: Partial occlusion drawing in autistic individuals has received little attention in previous research. In normally developing children field independence is positively correlated with partial occlusion drawing (Morra et al., 1996). The theory of Weak Central Coherence (WCC) posits that high field independence characterises autistic individuals. Thus the current study aimed to evaluate whether partial occlusion drawing represents an islet of ability for children with autism. METHOD: Autistic (N=21) and normally developing children (N=30) aged from 5;6 to 10;0 completed repeated measures of partial occlusion drawing and picture selection from two 3D models. Field independence was measured using the Children’s Embedded Figures Test. RESULTS: Logistic regression found chronological age, autism diagnosis and frequency of selection of partial occlusion to be significant predictors of partial occlusion drawing across all participants. Autistic children were significantly delayed in their development of partial occlusion drawing, selection of partial occlusion depictions as the ‘best’ did not differ significantly from controls. Autism was found to negatively predict partial occlusion drawing over that predicted by mental age. Field independence significantly correlated with partial occlusion drawing only in normally developing participants: the correlation in autistic participants was non-significant. DISCUSSION: The results speak against the WCC model of autism. That selection in autistic participants is normal but production is severely delayed suggests that there is a deficit in processes required to translate the intention to produce partial occlusion into a motor plan and to follow this plan to completion. The results are discussed within an Executive Function framework and it is suggested that deficits in working memory offer the most congruous explanation of the findings.