The study investigates term extraction methods using comparable corpora for interpreters. Simultaneous interpreting requires efficient use of highly specialised domain-specific terminology in the working languages of an interpreter with limited time to prepare for new topics. We evaluate several terminology extraction methods for Chinese and English using settings which replicate real-life scenarios, concerning the task difficulty, the range of terms and the amount of materials available, etc. We also investigate interpreters’ perception on the usefulness of automatic termlists. The results show the accuracy of the terminology extraction pipelines is not perfect, as their precision ranges from 27% on short texts to 83% on longer corpora for English, 24% to 31% on Chinese. Nevertheless, the use of even small corpora for specialised topics greatly facilitates interpreters in their preparation.