Many studies indicated that temperament and personality represent important categories of animal behavior. However, the description of animal individuality, encompassing broad and stable behavioral traits that generalize across time and situations, involves several difficulties. For example, there is often confusion between domain-specificity of behavior and generality and cross-situational consistency of individual differences. Adaptation dictates that behavior in general must be domain-specific, which would apparently make generalized individual differences impossible. However, applying concepts of behavioral consistency and generalizability from the human personality field can easily resolve this issue. "Consistency" is a correlational construct, meaning that the behavioral variable shows a high correlation over time or across situations even if its overall level changes. That is, an individual which is more fearful than others in one situation is likely to be so in another situation, even though the behavior overall differs in these situations.

When individuals, situations, response classes and times represent the basic conceptual units for the study of animal temperament, it becomes possible to define various kinds of behavioral consistency. For example, it can be possible to distinguish consistency of response profiles across situations (e.g. an individual may be high on fearfulness but low on dominance consistently in various situations) or consistency of time effects across situations (e.g. an individual may become more fearful with age consistently in various situations). A modular, hierarchical model of animal temperament, involving, at various levels, both broad and more situation-specific behavioral traits, is outlined.

Evidence will be presented that individual differences in many behavioral domains are consistent over time and across situations. They can be organized into a small number of dimensions, which represent a form of a hierarchical organization. These dimensions could be meaningfully interpreted in motivational, affective and cognitive terms, and are remarkably similar across species, both in the overt structure, function and underlying physiological mechanisms. Various approaches for measuring animal temperament will also be discussed, involving both objective ethological measurement, batteries of behavioral tests, as well as subjective rating assessment. The possibility that human observers may project the Big Five personality factors onto animal subjects will be also considered. Such possibility would in part explain why studies of animal personality based on observers' ratings usually reveal behavioral dimensions very similar to the Big Five personality factors found in humans.