Despite the growing importance of information technology (IT) interruptions for individual work, very little is known about their nature and consequences. This paper develops a taxonomy that classifies interruptions based on the relevance and structure of their content, and propositions that relate different interruption types to individual performance. A qualitative approach combining the use of log diaries of professional workers and semi-structured interviews with product development workers provide a preliminary validation of the taxonomy and propositions and allow for the discovery of a continuum of interruption events that fall in-between the extreme types in the taxonomy. The results show that some IT interruptions have positive effects on individual performance, whilst others have negative effects, or both. The taxonomy developed in the paper allows for a better understanding of the nature of different types of IT interruption and their consequences on individual work. By showing that different types of interruptions have different effects, the paper helps to explain and shed light on the inconsistent results of past research.