Beginning with a preconceived bias that "real" (i.e. nonartifactual) age differences in transfer and retroaction do exist, the author feels that the available literature permits no clear conclusions relating the process of aging and transfer mechanisms, or aging and retroaction. Research to date is viewed as assuming that "interference" manifests its effects over the age continuum and implies that an increased susceptibility to interference in the aged is unmodifiable. A more optimum research strategy, it is held, would concentrate on experimental manipulations which lead to the modification of age-performance functions. The bulk of the paper focuses on the research and implications which follow the adoption of this "modification" strategy. Three major classes of variables, of major interest in the study of learning and memory in the aged, provide the context: (1) degree of learning; (2) mediated vs. non-mediated learning; and (3) response time. The emphasis is on the provisions of alternate, empirical methods to provide for sources of confounding which mitigate against unbaised age-comparisons of the processes of retention and memory. (TL)