In this paper the question is addressed whether the word-level intelligibility of time-compressed speech can be improved over linear compression by using a type of non-linear compression. Two options are tested: one type of compression which takes into account the natural timing of fast speech; and one other type of compression that saves the segmental intelligibility of short unstressed syllables. This is tested at two rates of speech: fast and very fast. The results of the perception experiments (an articulation test and a speech-interference test) show that, at both rates of speech, neither of the two types of non-linear time-compression improves intelligibility over linear compression. This suggests that both the prosodic pattern and the segmental intelligibility of both syllables contribute to word recognition in fast speech.