The 'rule of thumb' that large volumes of water can be sampled for trace organic pollutants by XAD resin columns which are designed by small column laboratory studies or pure compounds is examined and shown to be a problem. A theory of multicomponent breakthrough is presented as a frame of reference to help solve the problem and develop useable criteria to aid the design of resin columns. An important part of the theory is the effect of humic substances on the breakthrough character of multicomponent chemical systems.