The evidence supporting the presence of output losses associated with sovereign defaults is based on annual observations and suffers from measurement and identification problems. This paper examines the impact of default on growth using quarterly data and finds that output contractions precede defaults and that output starts growing after the quarter in which the default took place. This indicates that default episodes mark the beginning of the economic recovery and that the negative effects of a default on output are likely to be driven by the anticipation of default, independently of whether or not the country ultimately decides to validate it.