This paper compares and contrasts the underlying principles of two conceptual systems—the interpersonal circle and the interpersonal octagon—for classifying and measuring a person's interpersonal tendencies. Both systems have been represented by two intersecting axes: a horizontal one extending from close/warm involvement to distant/cold separation and a vertical one extending from upper/control to lower/submission. In both systems, intermediate axes have been inserted between these two main ones. Where the circle would appear to be concerned with traits, the octagon is concerned with what have been termed states of relatedness. The two systems differ in their explanation and definition of adaptive and maladaptive relating behaviour. Whereas the circle has been closely aligned with the establishment of a bipolar relationship between the poles of the axes and with the mathematical model that is called the circumplex, the octagon has not. The most widely used circle‐based measure is the circumplex version of the Inventory of Interpersonal Problems. It generates high positive correlations between scales and a large, first general factor, but these imperfections have been corrected by the statistical procedure called ipsatization. The principal octagon‐based measure is the Person's Relating to Others Questionnaire. A number of high positive correlations have been demonstrated between the scales of the Inventory of Interpersonal Problems and the corresponding scales of the Person's Relating to Others Questionnaire. Therefore, despite there being differences in the underlying theories, the two questionnaires would appear to be measuring similar constructs. Copyright © 2012 John Wiley & Sons, Ltd.Key Practitioner Message: The interpersonal circle and the interpersonal octagon are two separate theoretical systems for representing a person's relating tendencies. Both systems are constructed around a horizontal, close‐distant axis and a vertical, upper‐lower axis. Both systems acknowledge that relating can be either adaptive or maladaptive, but where the circle incorporates both adaptive and maladaptive forms of relating within the same theoretical structure, there is a separate, adaptive (positive) octagon and and a separate, maladaptive (negative) octagon. Both systems have a questionnaire for the measurement of a person's general relating tendencies, but the octagon-based questionnaire is designed specifically to measure maladaptive (negative) relating. Interrelating concerns the relating that occurs between two specified individuals. Questionnaires have been developed by which each individual can record how, within the framework of the octagon, s/he relates (negatively) to the other and how s/he considers that the other relates (negatively) to her/him. There is no interrelating measure that is based upon the circle.