We examined the relationship between spontaneous gesture production and spoken lexical ability in children with Down syndrome (DS) in a naming task. Fifteen children with DS (3;8–8;3 years) were compared to 15 typically developing (TD) children matched for developmental age (DATD) (2;6–4;3 years of chronological age) and 15 matched for lexical ability identified by the MacArthur-Bates CDI questionnaire (LATD) (1;9–2;6 years of chronological age). Children of the DATD group displayed a larger number of correct spoken answers compared to other groups, while DS and LATD groups showed a similar naming accuracy. In comparison to both groups of TD children, a higher number of unintelligible answers was produced by children with DS, indicating that their spoken language is characterized by serious phono-articulatory difficulties. Although children with DS did not differ from DATD and LATD controls on the total number of gestures, they produced a significantly higher percentage of representational gestures. Furthermore, DATD children produced more spoken answers without gestures, LATD children produced more bimodal answers, while children with DS gestured more without speech. Results suggest that representational gestures may serve to express meanings when children's cognitive abilities outstrip their productive spoken language skills.