Traditional assumptions on processing of complex sentences can neither be studied separately in English, nor be directly applied to typologically different languages. The authors study focussed on (i) specific difficulties related to structural factors crosslinguistically, (ii) differences between children with and without language disorder and (iii) the correspondence between comprehension performance and general information processing abilities. The authors tested 12 school-age children with language disorder and 12 typically developing children matched individually on receptive vocabulary level. They examined the comprehension of different relative clause constructions in an act out task with toy animals. The statistic analysis revealed significant group differences and structural factors, i.e. interruption of main clause, head roles and perspective change. The comprehension performance of the sentence types showed the OS greater than SS greater than OO greater than SO order. Since the very same structures caused difficulties for both groups, the results support the non-representational accounts of language disorder. Including backward digit span as a covariate in the analysis eliminated all structural effects, and resulted in a weaker, but still significant group difference. This might show that many of the structural differences are in fact differences in the extent to which different structures charge working memory and executive functions.