Objective of this study was to elaborate and test hypotheses on the role of life stress in the onset of anxiety and depression. More specifically, it addressed the problem of the threshold value of life events (how threatening does an event have to be to be- come etiologically significant) and the role of long term difficulties (do they act as provoking agents or as vulnerability factors). Both aspects have been studied separately for the onset of anxiety, depression and mixed anxiety/depression. The results, obtained by means of a case-control design and the contextual interview and rating procedures of Brown and Harris, show that the onset of depression is preceded by severe events, the onset of anxiety by severe and moderately severe events, while the onset of mixed anxiety/depression was not related to the occurrence of events. The distinction between recently (within two years from onset) arisen difficulties and difficulties with a more distant onset appeared to be important: recent difficulties act as provoking agents for all diagnostic groups, whereas distant difficulties had no etiological meaning, not as provoking agent nor as vulnerability factor. There appeared to be no interaction between life events and recent difficulties, their effects were however additive. We conclude that life stress is implicated in the onset of anxiety, depression and mixed anxiety/depression and that recent difficulties are perhaps more important as antecedents than negative life events.