As is often remarked, in the first centuries of Islam, after the conquest of Iran, historians working in Arabic were in large measure influenced by Sasanian historiography, especially the late-Sasanian Pahlavi history known as the Khuday-namag. Given the Sasanian sources' own limitations, there are limitations to what one can learn, even with regard to the Sasanians themselves. Still, when modern scholars seek to address questions about pre- and early Islamic Iran, they turn to the Arabic sources. Many studies have enumerated the positive fruits of these labors. Surprisingly, even though much effort has been expended trying to see what the early Arabic sources can tell us about pre- and early Islamic Iran, virtually no systematic attention has been given to the other side of the question: What is it that the early Arabic historical tradition can't - or won't - reveal about pre- and early Islamic Iran, and whyn In this paper, I address this question through the case of the Sasanian capital of Ctesiphon and its topography. After its conquest in the 630s, Ctesiphon and its environs briefly served the Arabs as an administrative centre, but declined significantly. As early as the ninth century, detailed knowledge about Ctesiphon's topography was forgotten. When was Ctesiphon founded, and by whom Which monuments were located in Ctesiphon, versus in other of the cities (Arabic, mada'in)i That the Arabs referred to the entirety of the Sasanian metropolis, of which Ctesiphon represented one piece, as Mada'in, only added to the confusion (and which cities were the mada'inw). The impressive detritus of al-Mada'in demanded explanation, but with the death of generations who knew it in Sasanian times, left few clues. The paper's most important contribution is its application of the concepts of memory and forgetting to the Arabic sources regarding Ctesiphon. The sources include geographies, histories, and belles-lettristic texts, as well as core sources of Muslim religious tradition. The concepts of memory and forgetting draw attention to the power of texts to affect individuals and collectivities, and to create knowledge fundamental to all cultural, social and political interactions. Applying them to Arabic texts extends earlier scholarship on the literary dimensions of Arabic texts and the construction of historical meaning, now drawing attention to the power of texts to affect readers and listeners. Such knowledge is fundamental for group identities, and was critical in the early years of Islam in Iran.