Recent studies in speech production have reported articulatory strengthening and coarticulatory reduction of supralaryngeal gestures that gradually follow the hierarchical level of prominence or boundary in the prosodic structure of the utterance. The work presented here concerns the effect of prosodic boundary on linguopalatal articulatory dynamics in French, observed according to a four-level hierarchy of prosodic boundary: unaccented-syllable or unaccented-word boundary < accentual boundary < non terminal (“continuative”) intonational boundary < terminal (“conclusive”) intonational boundary. They were placed in median position in aC#Ca sequences including four types of consonant clusters: /kl/, /kt/, /lk/ and /tk/. Only 10 % of the 712 electropalatographic measurements done on the V and C segments and VC, CV and CC portions of the sequences show an articulatory co-variation increasing inter-individually (for at least two of the three speakers) and gradually according to the hierarchical level of the prosodic boundary. The stronger the prosodic boundary between constituents is, (i) the more the final vowel in the constituent is strengthened by an increase of its overall duration and the amplitude of its linguopalatal opening; (ii) the more the articulatory cohesion inside the rime of the final syllable is increased by a closer inter-gestural coordination between the end of the vowel and the beginning of the coda consonant; (iii) the more the boundary demarcation is strengthened by a reduction of the coproduction in the consonant cluster (but depending on its nature) and by a larger articulatory timing between the pre-boundary coda consonant and the vowel of the post-boundary CV syllable. These articulatory correlates of the prosodic structure essentially account for two or three hierarchical levels of the prosodic boundary. The supralaryngeal co-variations reflect less and less systematically the prosodic distinction in the following order: (i) first, intonational-group boundary vs accentual-group or word/syllable boundaries; (ii) second, accentual-group boundary vs word/syllable boundaries; (iii) then, terminal intonational boundary vs non terminal intonational boundary. These results inform on the depth of the architecture of the phonological representation of the hierarchical prosodic structure of the utterances, and on the phonetic nature of the relationships between supralaryngeal and suprasegmental dimensions of speech.