This paper describes a 2-year experiment on the flux with aging in A/Jax female mice of Arthus and tuberculin-type delayed hypersensitivities to chicken conalbumin antigen, of delayed hypersensitivity to methylated human serum albumin antigen, of anamnestic responses for these phenomena, and of a low-dose tolerance. The mice began to lose ability to develop delayed hypersensitivity at 15–18 months of age, or 0.6–0.7 mean life span. They retained ability to develop Arthus hypersensitivity, general anamnestic responsiveness, a capacity to develop low-dose tolerance preventing primary induction of delayed hypersensitivity and, once sensitized, the ability to express delayed hypersensitivity reactions well into senescence ( > 0.8 mean life span). These findings agree generally with most others on age-related flux of immunologic responses in mice and man, that cell-mediated responses are shorter-lived than humoral responses, and that induction (primary responsiveness) is much more affected than reaction or anamnesis (secondary responsiveness). Retention with aging of a capacity to develop tolerance, apparently not previously studied, could be important in explaining flagging primary responsiveness, because this (i.e. induction) is what such tolerance suppresses. Thus, certain unanticipated aspects of the experiment reported here suggest that accumulated exposure to food antigens may have contributed to age-related (i.e. time-related) decline of immunologic responses by progressive induction of such tolerance. The simple experimental model described here should be useful for investigating this possibility and the interplay of positive and negative immunologic responses with particular relevance to understanding such effects in aging human beings, since it studies natural changes in whole mice but with greater control, speed and dissectability that could ever be possible in man.