Purpose: The purpose of this study is to use prosodic and syllable-structure variation to probe the underlying representation of laryngeal kinematics in languages traditionally considered to differ in voicing typology (German vs. Dutch and French). Method: Transillumination and videofiberendoscopic filming were used to investigate the devoicing gesture in German, Dutch, and French for material that compared, first, a strong versus weak prosodic condition and, second, singletons versus clusters (stop + /r/ and /l/). Results: The results showed strengthening of the devoicing gesture in the strong prosodic condition and in the segmental context stop + /r/ for German and French but not for Dutch. In terms of timing (duration of oral occlusion, voice onset time, timing of peak glottal opening relative to stop release), French was intermediate between German and Dutch. Conclusions: (a) The representation of French voiceless plosives requires an active specification for glottal spreading just as in German; and (b) Static features are not well suited to capturing cross-language differences in voicing typology and changes in voicing specification over time.