This study examines the acquisition of the English laryngeal system by native speakers of (Belgian) Dutch. Both languages have a two-way laryngeal system, but while Dutch contrasts prevoiced with short-lag stops, English has a contrast between short-lag and long-lag stops. The primary aim of the article is to test two hypotheses on the acquisition process based on first language acquisition research: (1) native speakers of a voicing language will succeed in producing short-lag stops in the target aspirating language, since short-lag stops occur early in first language acquisition and can be considered unmarked and since one member of the contrast is formed by short-lag stops in both voicing and aspirating languages, and (2) native speakers of a voicing language will succeed in acquiring long-lag stops in the target language, because aspiration is an acoustically salient realization. The analysis is based on an examination of natural speech data (conversations between dyads of informants), combined with the results of a controlled reading task. Both types of data were gathered in Dutch as well as in EngDutch (i.e. the English speech of native speakers of Dutch). The analysis revealed an interesting pattern: while the first language (L1) Dutch speakers were successful in acquiring long-lag aspirated stops (confirming hypothesis 2), they did not acquire English short-lag stops (rejecting hypothesis 1). Instead of the target short-lag stops, the L1 Dutch speakers produced prevoiced stops and frequently transferred regressive voice assimilation with voiced stops as triggers from Dutch into English. Various explanations for this pattern in terms of acoustic salience, perceptual cues and training will be considered. (Contains 1 table, 3 figures, and 8 footnotes.)