Accuracy of our system, Contextualized Translation (CT), using the data set PhrasalVerbsToSpanish, together with the scores obtained with state-of-the-art machine translators: DeepL, Google Translator, Bing, and Yandex (all consulted in December 2017). Four baseline methods, based on looking up our bilingual lexicon of phrasal verbs, are also evaluated along with an unsupervised machine translation system, UNdreaMT (Artetxe et al. 2017). The last column shows the statistical significance test comparing each system with the previous one in the table.