This paper elaborates upon the observation, frequent in histories of human accomplishment, that fortuitous accidents play a role in discovery and invention. We present evidence from interviews with artists that suggests that accident is generally important to innovation process. We derive implications for business firms from this hypothesis. The interview data portray a process that incorporates accident and contrasts with the process representations and prescriptions of management researchers and other scientists. We consider the possibility that models in which accident plays a larger role (that is, models which more closely align with the process representations of artists) are more accurate and less idealized, and we discuss implications for managers of this possibility.