Interrogation in war and conflict is part of the wider system of military intelligence, the so-called human intelligence, which attracts the interests of historians and social scientists mostly to assess its effectiveness. Interrogation, however, is also part of the prisoner of war (POW) experience and it is a key one, because it often takes place right after the moment of capture and it represents the first close encounter with the enemy. At this moment, the “soldier hero” (G. Dawson, Soldier Heroes) is in an in-between zone, when he is not yet in the feminine world of passivity and captivity, and he can still fight, by withholding information. This paper focuses on POWs and their experience of captivity, looking at interrogation and questioning not from the point of view of its relevance to military intelligence and security studies, but taking into consideration the impact of cultural issues on the memories of these events. The POWs' heterogeneity and their diversity (for example being part of a minority or a racially distinct group) play a key role in the way the experience of captivity is remembered. Drawing on oral history interviews with WWII veterans from three racially distinct groups (African-American soldiers, Japanese-American soldiers and Native-American soldiers), the paper tries to show the depth of the impact of this heterogeneity on “memory composure” (A. Thomson, Anzac Memories). These veterans were also part of three groups which have become “legends” in the public memory narrative of WWII: the Tuskegee Airmen, the Nisei soldiers and the Navajo code talkers. Joining the wider debate on the struggle for racial justice, the paper takes into account the impact of these “legends” on the way these informants compose their memory, to conclude that the experience of captivity and interrogation is indeed remembered differently on account of racial heterogeneity. The paper is part of a wider research on interrogation and questioning in WWII which received funding by the Kluge Center (Library of Congress), Washington, D.C., and it is based on sources from the Veterans History Project collections at the Library of Congress, Washington, D.C. and the National Archives and Records Administration, College Park, MD.