Handling data imperfection in events is a critical issue in many application domains ranging from journalism and social media to healthcare communication and financial forecasting. In fact, failure to handle these imperfections can have significant consequences and lead to incorrect analysis and decision-making. In the literature, several typologies of data imperfections are proposed. However, these typologies can not be applied to events due to their particular features which typically include a trigger, arguments, argument roles, and temporal/spatial information. This paper introduces a typology of data imperfection in events, in a two-step process: (1), data imperfection types are identified based on an only single event mention. (2), the arguments of the same event are compared across multiple event mentions to further investigate additional imperfection types. This typology distinguishes among imperfection types that can impact the event’s entirety, types that specifically influence individual arguments within an event, and types that have the potential to affect both the complete occurrence of the event and its individual arguments. We finish by representing an illustrative example from an ontology- based memory prosthesis, handling imperfect events.