Part A of this paper is an analysis of the intermodulation method of distortion measurement. Results obtained by its use are compared with those obtained by the harmonic-measurement method. Predicted values for the intermodulation distortion and harmonic distortion are given for several typical transfer characteristics. For single-ended and push-pull characteristics, which are representable by simple power series, general equations are derived for intermodulation and harmonic distortion. With the aid of equations for the former, the effects of the ratio of signal amplitudes used in intermodulation testing are studied. These equations also permit derivation of relatively fixed ratios of per cent intermodulation distortion to per cent harmonic distortion for an intermodulation test method, which is described. Predicted values for distortion and their ratios are supported by test results. Curves expressing the actual distortion ratios, plotted against harmonic distortion, summarize the results of this analysis. These curves are useful for correlating the results of the two methods of test. Possible meter types, usable for metering the carrier- and intermodulation-frequency components in the intermodulation test method, are reviewed. The choice of meter type is found to affect the readings obtained for these components, and hence will affect the per cent intermodulation distortion. In Part B, simple equations are given for approximate predetermination of per cent intermodulation distortion from three or five points on the transfer characteristic. For more accurate prediction, tables are given for calculation of the prominent intermodulation components from eleven points on the transfer characteristics.