The most typical or prominent POS-grams, i.e., sequences of parts of speech or possibly other grammatical categories, can reveal a lot about the character of a text, especially with regard to its dynamics (reflected in the dominance of nominal or verbal constructions) or lexical density (the accumulation of lexical words as opposed to grammatical word sequences).In the study of translated Czech, previous research has shown that the POS-grams salient in translated texts differ from those in comparable non-translated Czech texts: they include more verbal combinations and pronouns. Their concrete realizations, e.g., the most frequent n-grams (sequences of n words) in given combinations, have indicated a possible interference effect based on the most represented source language: English.This study builds on the previous POS-gram and n-gram research on translated Czech and strives to describe and interpret the prominent POS-grams in translated Czech in the light of their corresponding English source texts, using a parallel corpus (namely, the English–Czech part of the InterCorp corpus). As a theoretical basis for description, hypotheses about translation universals are discussed. The results of the analysis indicate that some of the presumably universal translation tendencies can certainly be traced in Czech translations; however, translators' choices tend to be the result of a combination of factors rather than a single reason (such as explicitation or normalization). The study also comments on the specificities of cross-linguistic comparison based on POS-grams and n-grams in two typologically different languages.