Sometimes people laugh when interacting with social robots. Why? What are the laughable source and the sequential organization? What does the laugh reveal about Human-Robot-Interaction? This chapter is based on multimodal conversation analysis of video data from a setting in which participants naturally interact with the humanoid robot Pepper in a Danish institutional context. The chapter explores three variations of laughables: 1) laughs in response to conversational misfit; 2) laughs in response to script misfit; and 3) laughs in response to joke-telling. The chapter unpacks the basic sequential structure across these types. It discusses how laughs are produced as a consequence of the type of social setting in which bystanders are engaged in the participation framework. The chapter contributes to new understandings about how social robots should be able to engage in laugh sequences.