Abstract Clear speech describes a speaking style used in non-noisy conditions with the intent of maximizing clarity. The intelligibility advantage of clear speech has been shown to be linked to a reduction in speaking rate associated with the hyper- articulation of the style. This reduction in speaking rate involves numerous cues that carry multi-level (e.g. linguistic, perceptual, phonetic, acoustic) significance. Focusing only on the acoustic level, compared to its casual counterpart, clear speech exhibits an increase in pause frequency, pause duration and word duration. In pre- vious works, attempts to time-scale casual speech to match the segmental durations of clear speech have failed to increase intelligibility, due mainly to artefacts arising from the proposed modifications and the absence of pauses in the hypo-articulated casual speech. Consequently, this work explores alternative time-scaling algorithms that mimic both elongations and pauses observed in clear speech, while respecting the acoustics of the unmodified casual speech in an attempt to avoid disturbing arte- facts. First, uniform time-scaling is employed as a baseline modification so that the casual speech sentence duration matches that of the clear. Then, time-scaling ap- proaches are examined that seek to both elongate casual speech and insert pauses. Specifically, two refined approaches are proposed that exploit acoustic properties of the speech, namely loudness and stationarity, in determining where and how to elon- gate certain parts of the signal and insert pauses. Evaluations of the time-scaling approaches were conducted via listening tests considering speech-in-noise intelligi- bility. While the uniform time-scaling achieved essentially the same intelligibility as the unmodified casual speech, with some improvement observed for non-native listeners, the more refined approaches with elongations and pause insertions did not prove advantageous. These results would suggest more careful consideration of time-scaling approaches aiming to decrease speaking rate and increase intelligibil- ity, potentially requiring incorporation of more linguistic information in multi-level acoustic-phonetic analyses.