the objective of this study was to evaluate the effect of the rainfall variability on the runoff and sediment amount in the pipiripau river basin (a = 235 km2), in the distrito federal, brazil. to represent the climatic variability of the historic precipitation series, 5 years were selected: the moistest, the driest, an average, and 2 years representing the standard deviations from the series (half-wet and half-dry). after the calibration of the model with basin rainfall and runoff data, each rainfall event was simulated in each one of the selected years, and sediment yields were computed using software musle for three different land-use and management situations: (a) the present conditions, with predominantly conventional agriculture in the agricultural areas (pc); (b) the former scenario, where native (cerrado) vegetation existed in the basin (cer); and (c) no-till in the agricultural areas (pd). although the relationship of the mean annual flow rates with the annual rainfall volume was linear, there was an exponential increase in runoff volume and sediment yield with precipitation in all scenarios; the most significant increases were observed for pc, followed by pd and cer. the exponential increase can be explained by the non-linearity of the musle model regarding both precipitation and runoff volume.