This paper provides acoustic correlates of intonation in whispered, semi-whispered and normal speech modes. In particular, it investigates correlates of utterance-final rising intonation in polar questions and falling intonation in statements. The paper does not only examine properties of vowels but also properties of the following voiceless consonant clusters. For the purpose of this study 2592 items produced by 16 native speakers of Polish were analysed. The results point to differences in spectral properties of both utterance-final vowels and consonants where falling intonation in statements contrasts with rising intonation in polar questions. Regarding the consonants, questions are produced with higher peaks, intensity, COG and STD values as well as smaller skewness and kurtosis values. Some spectral differences of consonants, including spectral slopes, are more distinguishable for questions versus statements in the whispered speech mode than in other speech modes. The more pronounced role of these cues in whispered speech suggests their compensatory function for the fundamental frequency, which is present in phonated speech. In summary, the study shows that speakers produce intended intonation patterns by varying the choice of cues as well as their magnitude in dependence on both (i) speech modes and (ii) intonation patterns.