Second language learners must acquire the ability to use word boundary cues to segment continuous speech into meaningful words. Previous studies have used two types of s+stop clusters to test second language English speakers on their ability to segment fluent English speech: cross-boundary clusters (this table) where allophonic aspiration is present and word-initial clusters (this stable) where allophonic aspiration is absent. These studies suggested that first language segmentation strategies influence second language segmentation. The goal of this study was to test real-time processing of these cluster types by second language learners from one language where cue adaptation was possible (Mandarin Chinese) and one where a new cue would have to be learned (French). Results did not support the idea that first language segmentation strategies influence second language segmentation, but found that both language groups had high accuracy of identification despite showing uncertainty in real-time processing.