Keywords: 
• rainfall-runoff relationship /
• hydrograph separation /
• experimental basin /
• environmental isotope

Abstract: Empirical graphic separation methods with large uncertainties are routinely used for decades to obtain input data,e.g.,the unit hydrograph etc,for various hydrologic models of rainfall-runoff.During 1970's another tracer-based hydrograph separation which attempts to be more physically-based than the above method mostly involves a two-component mixing model for stream flow which relies on the simple mass balance equations of water and isotope is widely used abroad.Eight fundamental assumptions involved in this method and its reality are tested within experimental and representative basins since 1987.From thin work,it is concluded that due to the isotopic situation is much more complex than these assumptions envisioned,most of them seems to be unrealistic notions,it follows that the conclusions of most previous studies based on this kind of two-component mixing model e.g.,a much higher propotion of storm runoff is contributed by groundwater,should be reconsidered.It is also concluded from this work that our reyuisities towards appropriate hydrograph separation into two or more than two components natural basins should be considered.