In the present study we look into Serbian seven-year-olds’ understanding of literally translated English idioms in order to determine whether inherent visuo-spatial information facilitates the interpretation process. Drawing on our previous research involving adult respondents (Antović and Stamenković 2012), the present report tests the degree to which the existence of lexicalized visual and spatial configurations in unknown idiomatic expressions aids their understanding with seven-year-olds. We have presented 125 Serbian children with 6 literally translated English idioms containing lexicalized visual and spatial configurations (e.g. ‘put the cat among the pigeons’) and 6 literally translated English idioms with no visual or spatial component (e.g. ‘have a sweet tooth’). For each idiom, the children respondents had the task of circling a letter beside one of the four given drawings which they thought best described the meaning of the idiom in question. The idioms were randomly selected among high-frequency expressions with no direct equivalents in Serbian, available in Collins Cobuild Dictionary of Idioms (1995), Cambridge International Dictionary of Idioms (1998) and Oxford Dictionary of Idioms (1999). The preliminary result suggests a considerably better understanding of idioms with visual and spatial configurations (59%), as compared with those having no visual or spatial component (38%). Along with our previous results with adults, this might provide some more support to theoretical claims presented by authors such as Jean Mandler - that there is a spatial foundation of the conceptual system.