This paper addresses the question to what extent the life story reflects the narrator's self-coherence. In order to provide an answer, we relate various measures of remembering and narrative coherence which we assume to measure self-coherence. Our approach takes as a starting point Fitzgerald's self-narrative theory (1992, 1996), which is aimed at explaining the reminiscence-bump phenomenon by conceiving of personal coherence (integrated identity) as reflected in the way recalled significant life events are distributed in autobiographical time. Following Fitzgerald's (1992) method, we examine data obtained from 25 elderly persons. The results clearly confirm the reminiscence-bump reported earlier: the relative frequency of stories increase for the 16-20 years age period, reach a significant maximum in the age from 21 to 30 years, and decline radically in the age from 31 to 35. In order to measure narrative coherence we analyze the autobiographical texts according to Gergen and Gergen's (1988) criteria construed to assess well-formedness of personal narratives. The application of this measure yields a reminiscence-bump which is restricted to a narrower range, the period of 21 to 25 years of age. Our findings can be taken to elaborate on the claims of the self-narrative theory in that: 1. persons with the most well-formed and most coherent life stories select their significant memories predominantly from a rather narrow age period, yielding a reminiscence-bump; 2. in contrast, persons with less well-formed stories fail to attain a reminiscence-bump; 3. comparison of the chronological pattern and the narrative well-formedness of memories, independent of persons' well-formedness groups, provides additional support for the validity of a narrowed period for the reminiscence-bump. Finally, we apply Schank's (1990, 1999) method of story-indexing which is used for analyzing autobiographical narratives to identify the 'generalized lessons and beliefs' being reflected by stories of life events. Our analysis reveals no relationship between the reminiscence-bump phenomenon and story-indexing. Persons' tendency to recall coherent stories, however, is found to correlate with their tendency to elicit stories with higher values of indexability. These findings are interpreted to indicate that the explicit use of generalized lessons and beliefs in guiding the recall of significant life events is not confined to stories from a particular age, but it is a necessary prerequisite for the construction of a coherent life story.