An accurate assessment of cognitive abilities in populations that differ from the majority in cultural and linguistic characteristics is one of the main challenges in cognitive testing. Previously developed methods for assessment of the validity of cognitive scores in individuals with diverse backgrounds, such as the Culture-Language Interpretative Matrix (C-LIM), have not been empirically substantiated. We tested the applicability of the C-LIM in the European context, by comparing selected test scores from the Woodcock-Johnson-IV Test of Cognitive Abilities (WJ-IV) between Roma children aged 7-11 years (n = 399) and their counterparts from the normative population (n = 131). The largest differences were detected in WJ-IV tests requiring abstract reasoning and manipulation with complex signs. Furthermore, the C-LIM did not reliably discriminate between our groups and its use appears to be inappropriate for making diagnostic decisions about children from populations that do not traditionally rely on processes such as categorical thinking, abstract reasoning, and generalization.