Background: Individuals with developmental language disorder or Williams syndrome are reported to use more gestures than individuals with typical development. However, these two groups differ considerably in visuospatial and language skills, two skills that are hypothesized to shape gesture rate. Aims: We first examined whether children with both developmental language disorder and children Williams syndrome do indeed use more gestures. Our second aim was to disentangle the role of vocabulary and visuospatial skills in the use of supplementary gestures (i.e., containing unique information). To account for participant heterogeneity, analyses included both group comparisons and vocabulary and visuospatial skills at an individual level. As a third aim, the role of visuospatial skills was further examined in relation to gestures containing spatial content. Methods & Procedures: In a cross‐sectional group design, three participant groups watched and then retold a cartoon: children with typical development (n = 25), children with developmental language disorder (n = 25) and children/young people with Williams syndrome (n = 14). Their narrations were transcribed and hand gestures were coded based on gesture–speech integration (redundant, adds information to particular lexical items, gives information that is entirely absent from speech) and spatial content. Participants' expressive vocabulary and visuospatial skills were measured. Outcomes & Results: Between‐group comparisons showed that individuals with developmental language disorder or Williams syndrome did indeed use more gestures. Poisson loglinear modelling demonstrated that a relative higher use of supplementary gestures was determined by lower expressive vocabulary skills. Neither the group distinction nor visuospatial skills shaped the supplementary gesture rate nor spatial gesture rate. Conclusions & Implications: Regardless of neurodevelopmental condition or typical development, a higher use of supplementary gestures was influenced by expressive vocabulary skills. Children with lower vocabulary skills spontaneously capitalized on the multimodality of communication to express constituents that were not present in their verbal speech. This finding is a promising starting point for future gesture intervention studies examining whether implicit modelling of gesture use can encourage gestures even more in these children and if this allows them to achieve higher linguistic complexity. On a methodological note, the observed intra‐group skill variability demonstrates that group comparisons need to be complemented with correlational measures accounting for skills at an individual level. What this paper adds: What is already known on the subject: Children with developmental language disorder and children with Williams syndrome are more inclined to use gestures than typically developing children. Research conducted in adults with typical development points towards the role of lexical and visuospatial skills in gesture use, but it is unclear how these skills shape gesture use in children with atypical development. What this paper adds to existing knowledge: This study compares the rate of gestures that convey meaning that is not expressed in speech between the three aforementioned populations. Novel is the inclusion of the group distinction, individual lexical skills, and visuospatial skills in one encompassing statistical model. What are the potential or actual clinical implications of this work?: The inclination to use gestures that replace speech is related to lexical skills. Visuospatial skills do not seem to play a role and should not be considered as a factor when thinking about gesture intervention. Understanding how gestures relate to specific skills is a first step to understanding how gesture interventions can bolster language production. [ABSTRACT FROM AUTHOR]