The author has collected a database of several thousand of Holodomor oral history sources during preparation of GIS-Atlas of the Holodomor. It was this base that made possible comparison of similar or identical testimonies of the same or various respondents, published in numerous collection of Holodomor oral sources. The goal of the article: to figure out how a narrative of one person published in various editions is correlated, from the point of view of recalling of events during the Holodomor. Methodology: historical method, analysis and synthesis, methods of source analysis (comparative, textological). Scientific novelty. The author focuses not on verification of oral history sources by the documents or other oral sources, but on credibility of given data from the point of view of specificity of recalling and process of collection and preparation of oral history for publication. Conclusions. The research indicates that on the one hand, there is a certain percentage of republication of testimonies by various editors that, it seems, decrease source potential of an edition; on the other – such duplication of testimonies is often a valuable material for understanding of the fact how narratives have been changing with times, time frames of narratives; what events were the first to come to mind and were important for a person, how a person was recalling. The author argues that there is a certain core of an informant’s narrative that can be supplemented with the details significant for a respondent. It is revealed that information field does not influence core of narrative that contains vital for a certain informant memories, mostly tied with the struggle for life (searches, survival strategies) and defeat of this struggle (mortality). The article suggests differences in recalling of events by respondents during various years. The revealed problem of falsifying of testimonies is tied first of all with a figure of interviewer and is not a mass phenomenon.