This part of ISO 555 gives the principles of the integration method and describes several sampling procedures that may be used to establish that the tracer mixing is satisfactory at the cross-section where the measurement of flow is required. It deals with the choice of tracer and gives a brief outline of the channel characteristics necessary for the application of the method. The procedures that maybe required to inject the tracer, to determine the quantity injected and to measure the concentrations of tracer in the injection solution and in the samples are given. The complexity of these procedures can vary considerably, depending on the objectives. When only a rough estimate of discharge is required in a remote area or during spate flow conditions, the experimental and analytical techniques can be far less rigorous than those required when calibration of a gauging structure is required. Sample analysis is not dealt with in detail but a numerical example is given to illustrate the calculation of flow and the estimation of uncertainties.Field of applicationFor a given discharge, the integration method requires less tracer than the constant-rate injection method. It is, therefore, particularly applicable for high discharges or where tracer economy is important because of considerations of cost. Discharges of several thousand cubic metres per second have been determined using the integration method. In the case of smaller flows, the method may be preferred on sites where access is difficult or when facilities are sparse, because the field operations can be very simple. Should any doubt exist as to the steadiness of the flow, the constant-rate injection method should be used, since the integration method may provide no information on the variation of flow with time and, consequently, errors in the computed flow would occur. In narrow channels such as sewers where the tracer concentrations may rise and fall too rapidly for samples to be taken in such a way as to represent accurately the passage of tracer, the constant-rate injection method shall be used (see ISO 555-1).The method can only be applied when the mixing processes within the channel, whether caused by natural turbulence or otherwise, are sufficient for the tracer to be thoroughly mixed across the entire section of the channel at the sampling station and when all the tracer injected passes through the sampling cross-section. Dilution gauging is particularly suitable for turbulent rivers where other methods are difficult to apply, e.g. rock-strewn shallow rivers with steep slopes. The method may be used to calibrate other methods because dilution gauging involves absolute determinations of volumes and times and determinations of relative values of concentrations only. When calibrating other methods, the accuracy and cost of dilution gauging should be compatible with the device or structure under examination and for guidance on the correlation to be expected between results reference should be made to the relevant International Standards dealing with the other methods.