The paper presents the analysis of the humor found in four dyadic conversations. The results of the conversational data match those of previous studies (Pickering et al., 2009): no differences were found in volume or speech-rate between humorous pause units and non-humorous ones. Similarly, pauses were not found to mark humorous turns. However, the result that punch-lines showed lower pitch than non-humorous parts of the text was not replicated: humorous pause units showed no significant differences in pitch from non-humorous ones. Smiling is found to mark humor only in a general sense of “setting the frame” and is not integrated (i.e., co-extensive) with the humor.