ABSTRACT: In most approaches to modeling the rainfall/runoff process, a spatially lumped description of precipitation has been assumed adequate for modeling the important characteristics of catchment response. However, theories of catchment hydrology as well as some re cent modeling studies suggest that spatial variability in precipitation may be important in determining the characteristics of stream flow hydrographs. Data from two intensive rainfall recording experiments in Illinois have been used to examine the effects of rainfall pattern on stream hydrographs for summer convective storms. A threshold analysis was introduced to distinguish storms of markedly different pattern. A mixed deterministic/stochastic modeling procedure was used to generate hydrographs resulting from storms of different patterns. It was found that differences in peak timing resulting from different patterns were highly significant, differences in the distributions of peak flow were significant and differences in the distributions of storm flow volumes were insignificant even given a long period of synthetic record. Rainfall pattern per se seems to have only a second order effect on hydrograph characteristics. Assessing the correct volume of input in a highly spatial variable pattern is far more important in predicting stream hydrographs.