This paper examines the manner in which age is treated in the design and conception of developmental research, and the dissatisfaction with its status as a concept and its use in research. This is attributed to the widespread tendency to consider age as an independent variable, comparable to others employed in differential research, and to study age differences rather than age changes. An alternative view is presented, which treats developmental questions as analogous to other phenomena involving changes in behavior over time. Age is looked at as a dimension along which behavior changes are to be traced, forming part of the definition of the dependent variable in developmental studies. This dependent variable is to be defined in terms of specified aspects or parameters of the function describing the changes which occur with age for a given behavioral variable. Implications of this view for developmental research are brought out, with particular reference to problems of description and quantitative analysis of developmental change, to the application of the experimental method in developmental research, and to the study of individual differences in the context of developmental change.