Which information do people seek out when trying to explain everyday events? Previous research (Ahn et al., 1995)indicates that this may not be the same information that people take into account when provided, and that theories ofcausal reasoning consider crucial. In an experiment, we asked participants to generate questions to explain type or tokenevents, which were familiar or unfamiliar. Based on theories of singular causation, we expected participants to search forpresent causes and indicators of actual causation to explain token events, but for causes and their covariations with theeffect when explaining types of events. We assumed participants to inquire about the presence of known causes whenevents are familiar, but about potential causes when events are not familiar. We categorised generated questions accordingto the information sought. Results partially supported our predictions. We discuss the relevance of the findings for differenttheories of causal reasoning.