The general problem of the incorporation of IR compensation in a potentiostatic circuit is considered. The manner in which the number of time constants needed to describe the system affect the amount of stable positive feedback at short times or high frequencies is discussed and a theoretical model developed for present state of the art potentiostats and operational amplifiers. It is shown that the use of a model in which all series nonfaradaic resistance has been removed to establish a reference for complete IR compensation indicates that stable overcompensation cannot take place in any practical system. The use of positive feedback in pulse and a-c techniques is compared and it is concluded that great caution must be exercised in the latter with regard to the proper detection of optimum compensation. Interpretation of results using IR compensation is considered and it is shown that the use of frequency domain conversion to obtain the impedance of the system under, e.g., pulse conditions allows potentiostat parameters to be eliminated such that meaningful results may be obtained at times of the order of 10 nsec. © 1969, The Electrochemical Society, Inc. All rights reserved.