This research is carried out in Elo sub catchments area that contains of the regency and the municipality of Magelang and Temanggung. The goals of this study are: (1) to predict the peak discharge by using the characteristic variable of catchments physical environment by means of the aerial photo interpretation, to examine the technical accuracy of remote sensing to detect coeficient run off which used to predict the peak discharge. The research uses the aerial photo of black and white panchromatic on the scale 1:50. 000 in I994. It issupported by secondary data and field survey.The research method is the integration between the technical remote sensing and the Geographical Information System (SIG). Field survey is used to check the result of aerial photo interpretation and to collect the data that cannot he obtained from the image. The sample uses the purposive stratified sampling.The results of the research show that: (1) the aerial photo of black and white panchromatic on the scale 1:50. 000 is obtained to interpret the slope area and it is not good to interpret the accuracy of the covered arm is about 87% and 77,4 %, (2) the current coeflicient which is accounted hased on the table of Cook method is about 40,36% - deviates 39,94% from the field data; (3) the peak discharge in Elo sub catchments area with rational formula is about 3022,03 m/second, while the peak discharge as the observation result is about 112,675 m/second. Its large of research area causs the time to peak takes a long time, which causes raining intensity increases. The increasing value of raining intensity and the flow coeficient makes the value of peak discharge increases that larger than measured peak discharge.