In languages such as Japanese or Korean, most research on the expression of politeness focuses on morphological and lexical honorifics. Here, we ask whether listeners can perceive the intended honorific level of Korean utterances even in the absence of explicit verbal markers, and whether these phonetic cues are available cross-linguistically. In two perception experiments, we found that Korean listeners perceived intended honorific levels with above chance accuracy when listening to utterance fragments that gave no verbal cues for honorification. In Experiment 1, stimuli from multiple voices were presented at random and participants had to judge the intended honorific level of isolated stimuli. Overall accuracies were low (58% for Koreans; 53% for English listeners). In Experiment 2, we blocked the presentation of different voices and asked participants to compare honorific and non-honorific speech from the same voice. Accuracies increased to 70% for Koreans and 57% for English listeners, indicating that speech acoustics become an important cue for politeness-related meanings when listeners can compare utterances spoken by the same speaker. Our work shows that politeness does not merely reside in verbal markers alone but is co-signaled by phonetic cues. And, because the English listeners performed above chance level in Experiment 2, the results furthermore suggest that at least some acoustic correlates of politeness may be understood in a similar fashion across cultures.