Linguistic interference in simultaneous interpreting is among those phenomena that many authors have written about, while few have actually investigated it. Following Daniel Gile’s request for more empirical data, the authors have tried to analyse frequency and types of interference in a corpus of 36 interpretations by twelve professional conference interpreters. Results indicate the high incidence of interference (INT) in professional interpreters’ output as well as the high variability in both frequency and type of INT among the subjects. The lack of correlations between INT and other investigated parameters seems to indicate a certain independence of INT from other output parameters (e.g. semantic deviations).