The objective of this paper is to summarize approaches to translated texts in international corpus linguistics, to introduce the CNC corpus design regarding translations, and to present some examples which demonstrate that texts translated into Czech may differ from original Czech texts and affect the research results.