Temporal gradient (TG), i.e., differential recall of recent and old memories, is a well known feature of amnesia. A recent study provided evidence of a classical TG for media-mediated events in elderly healthy people, showing that they recall remote events significantly better than recent ones, while a reverse TG, i.e., better recall of more recent events, was demonstrated in younger normal subjects. In the present study we present normative data which, using the same test, allow evaluation of TG in single cases and their qualification as classical or reverse. The normative procedure was also applied to a small sample of subjects with probable Alzheimer's disease or mild cognitive impairment. Norms for TG may be helpful not only to assess healthy people's performance, but also to judge any apparent TG in pathological subjects.