The aim of the study was to demonstrate the potential of aligned texts in a parallel corpus for research. It has been prompted by the fact that, at the moment, a coordinated effort is being made to build a parallel Czech-English and English-Czech corpus at Charles University in Prague and Masaryk University in Brno. There is no doubt that parallel corpora provide a unique opportunity and a powerful impetus for contrastive investigation. The study focuses on the conspicuously little explored subject of the diverse uses of the Czech conditional, or rather its morphological exponent by (a special form of the Czech verb to be subsuming six inflected forms), and its correlates in English. An interesting corollary is a comparison of the translation equivalents found in the texts with the equivalents offered by the largest and most recent Czech-English dictionary Fronek 2002). As a matter of fact, the dictionary gives only two equivalents of the Czech conditional lexeme by, namely would and should (as the 1st person conditional and the central modal auxiliary of obligation, synonymous with ought to). The minicorpus of Czech-English texts is comprised of two parallel texts aligned by means of the Vanilla aligner: the original of a novel by the wellknown contemporary Czech writer Michael Viewegh and its English translation made by a British translator (see References). The Czech original consists of 42 822 tokens/13 834 types (WordSmith Tools), the translation (56 127 tokens/8 107 types) is by 31 per cent longer in tokens, a fact which is partly due to some of the periphrastic equivalents of the Czech by.