The behaviour of microorganisms in urban stormwater should be thoroughly investigated and understood to (a) design treatment technologies that can reduce the human health risks of utilising stormwater and (b) develop models which can accurately predict the levels of microorganisms in urban stormwater to aid in health risk assessments. A crucial part of understanding the behaviour of pollutants in urban stormwater is to determine whether the pollutant experiences higher levels in certain portions of the event (e.g. does the pollutant experience a first flush?). The aim of this paper is twofold: (a) determine if the first flush phenomenon exists for a commonly used microbial indicator, Escherichia coli, and (b) determine whether the presence of a first flush is dependent on antecedent climatic and/or hydrologic characteristics. E. coli data collected from the wet weather flows of four urban catchments in Melbourne was used in the paper. Cumulative mass versus volume curves were used in conjunction with standard statistical inferences to determine that the first flush phenomenon was not consistently present, and that the presence and magnitude of a first flush varied considerably between each site. Regression analyses were used to determine that this variation was probably not caused by the same governing processes for all four sites, with different explanatory variables significantly explaining the first flush at each site.