Background: Near miss events, defined as harm averted due to chance, are learning opportunities in radiation oncology. Psychological safety is a feature of a learning environment characterized by interpersonal risk taking. We examine the effects of near miss type and psychological safety on reporting near miss events to an incident learning system. We posit that submission likelihood will differ based on near miss types and psychological safety. Methods: We administered a survey assessing psychological safety to members of a radiation oncology department. We then presented six events for a patient with a pacemaker (PM), which requires cardiac clearance before radiation: process-based (harm averted by systematic PM check); good catch (harm averted by incidental PM check); “could” event (harm averted by chance PM absence); “almost” event (positive PM status, but no sequelae); hit (positive PM status, subsequent arrhythmia); and control (no PM, checked PM status, no sequelae). Subjects ranked each event on submission likelihood (1 = most likely, 7 = least likely), scored events based on submission likelihood by others (1 = least likely, 7 = most likely), and rated event success (1 = failure, 7 = success). ANOVA was used to assess differences in mean rank, submission likelihood, and success ratings. Regression was used to assess the relationship between psychological safety and submission likelihood. Results: The survey yielded 95 out of 127 responses (75%). Mean ranks (p < 0.0001), submission likelihood (p = .042), and success ratings (p < 0.001) differed by near miss type; psychological safety predicted likelihood of submitting the different near miss types (Table 1). 14 respondents (15%) would mind if an incident was submitted about them, while 43 respondents (45%) assume others would mind if an incident was reported related to the others (score > 4). Conclusions: Near miss events proximal to a negative outcome are more likely to be reported, though this effect may be mediated by psychological safety.