Three- and four-way conversations in background noise were analysed for the frequency and timing of head turns between other participants. In both situations, it was found that head turns between interlocutors were much more frequent when the individual was speaking, than when they were listening to exchanges between the others. In the four-way conversation, turns were more frequent with increasing sound level when listening, but were independent of sound level when talking. The overall frequency of head movements was also greater in the four-way conversations than in the three-way conversations, both when speaking and when listening. Moreover, the participants turned earlier, anticipating the exchange of floor more frequently than in the 3-way conversations. However, because these data come from different experiments, accounting for these differences is not straightforward. For instance, exchanges of conversational floor were roughly twice as frequent in the four-way conversation, which may explain the increase in head movement when listening, but not while speaking. The four-way conversations also manipulated hearing impairment and hearing-aid use. Hearing aids for the two hearing-impaired listeners in a conversation changed the behaviour of both groups with more frequent head movement for the impaired participants and less frequent movement for the unimpaired participants. Again, these effects were apparent both when a given participant was speaking and when listening to others.