Music is a major element of social gatherings. However, creating playlists that suit everyone’s tastes and the mood of the group can require a large amount of manual effort. To solve this problem we present MoodMusic, a method to dynamically generate contextually appropriate music playlists for groups of people. MoodMusic uses speaker pitch and intensity in the conversation to determine the current ‘mood.’ It then queries the online music libraries of the speakers to choose songs appropriate for that mood. This allows groups to listen to music appropriate for their current mood without managing playlists. This work contributes a novel method for dynamically creating music playlists for groups based on their music preferences and current mood.