Older people often complain of difficulties in understanding speech in noisy circumstances. The current study tested the hypothesis that problems segmenting speech may contribute to these difficulties. Segmentation ability was measured in young normal-hearing, older normal-hearing and older hearing-impaired listeners. Listeners were presented with sentences in competing speech and resultant misperceptions were analyzed in terms of their accordance with the metrical segmentation strategy. While strong support for this strategy was indicated, no difference in the use of this strategy was found across the three listener groups, suggesting older listeners were unlikely to be experiencing segmentation difficulties at the sub-lexical level.