A considerable body of evidence is presented showing a substantial degree of longitudinal as well as cross-situational consistency in the motive area of aggression. These data demonstrate that the influential conclusions with regard to consistency drawn by Mischel in his evaluative review (1968, 1969) are not supported by existing empirical evidence in the field of aggression. The consistency found makes it both defensible and natural to assume the 'existence' of some kind of relatively stable, individual-differentiating aggressive reaction tendencies within the individuals, however conceptualized. Furthermore, on the basis of a conceptual analysis, it is shown why low correlations have often been found between self-report data (S data) and rating data (R data), on one hand, and objectively recorded behaviour (T data), on the other. It is argued that the reasons are often to be found in a conceptual mismatch as well as in psychometric insufficiences of T data. This analysis has broad implications for personality measurement in laboratory settings in general.Several times in the history of personality psychology and maybe particularly during the last decade, there have been lively discussions and heated controversies about what has become known as the consistency issue. This problem includes both questions of cross-situational and longitudinal consistency. Broadly, the question of cross-situational consistency concerns the extent to which individuals in a group retain their relative positions on a certain dimension or characteristic across various situations or sources of data at approximately the same point in time. The issue of longitudinal consistency, on the other hand, primarily relates to the extent to which individuals in a group retain their relative positions on a certain dimension or characteristic for measurements at different periods of time, that is, it concerns the degree of stability of individual differences over time (Olweus, 1974(Olweus, , 1977~1, c, 1978.The issue of the extent to which there is consistency across situations and data sources and over time is of great interest in its own right, for instance, for predictive purposes. However, closely related to this complex of problems is another question : Is there empirical justification for assuming relatively stable reaction tendencies or personality dispositions that affect the individual's behaviour in a number of situations? An assumption of such personality dispositions in one form or another is central to traditional trait theories (Allport, 1937; Cattell, 1946).The present paper will consider some of these issues, particularly in the following two ways. In the first place, it will briefly document the position of Walter Mischel (1968, 1969) who has been one of the most ardent and articulate critics of trait (and state, i.e. psychodynamic) theories. After a statement of Mischel's position, an overview of the evidence on longitudinal and, to some extent, cross-situational consistency in one area of research, aggression, will be prese...