Guicang 歸藏 "Return to the storehouse" was an ancient text that is only preserved in fragments. It was one of three books of "changes" (Yi 易) and was therefore also known with the title of Kunqian 坤乾, which is a reversion of the two hexagrams qian 乾 and kun 坤 known from the Confucian Classic Yijing 易經. In the ritual Classic Zhouli 周禮 it is said that the royal Grand Diviner (taibu 太卜) mastered three different methods of divination, one of which was that of the famous Zhouyi 周易 (i.e. the Yijing), a second one the Guicang, and the third that of the Lianshan 連山.
The late Han-period 漢 (206 BCE-220 CE) scholar Zheng Xuan 鄭玄 (127-200) was of the opinion that the Guicang was the standard mantic book of the Shang period 商 (17th-11th cent. BCE), before King Wen 周文王 of the Zhou dynasty 周 (11th cent.-221 BCE) introduced the famous Yijing. The first hexagram of the Guicang, kun, is the symbol of the earth, the storehouse (cang 藏) to which all plants and small animals return (gui 歸) in autumn. The obverse sequence of the two hexagrams seems to reflect a matrilinear tendency of the ruling house of the Shang dynasty.
The Guicang is mentioned in the Tang-period 唐 (618-907) encyclopaedia Beitang shuchao 北堂書抄 and the Song-period 宋 (960-1279) encylopaedia Taiping yulan 太平御覽. Both quote a lost fragment of Huan Tan's 桓譚 (23 BCE-56 CE) Xinlun 新論, and it is also quoted by the Eastern-Jin-period 東晉 (317-420) master Guo Pu 郭璞 (276-324). It can thus be known that at least parts of the Guicang still existed during the Han period. It is, nevertheless, not mentioned in the imperial bibliography Yiwen zhi 藝文志 in the official dynastic history Hanshu 漢書.
Later versions of the text seem to be forgeries. The modern scholar Wang Guowei 王國維 (1877-1927) analysed a sentence quoted in the Taiping yulan and came to the conclusion that the Guicang was indeed a mantic book from the Shang period, as can be seen in comparison with oracle bone inscriptions from that time.
The author of the Guicang is not known, yet the now lost Jin-period 晉 (265-420) catalogue Zhongjingbu 中經簿 says that is was a very early text written before any Confucian writing was compiled. According to the imperial bibliography Jingji zhi 經籍志 in the history book Suishu 隋書, it had a length of 3 juan. There was a commentary written by the Jin-period scholar Xue Zhen 薛貞. Another commentary was written by Sima Ying 司馬膺 (507-577).
After the Tang period, the book was lost, barring the chapters Chujing 初經, Qimu 齊母 and Benshi 本蓍. Yet even these chapters were forgeries, as Ouyang Xiu 歐陽修 (1007-1072) assumed. The Yuan-period 元 (1279-1368) scholar Ma Duanlin 馬端臨 (1254-1323) called them concoctions by a certain Liu Xuan 劉炫.
Fragments of the Guicang were collected by the Qing-period 清 (1644-1911) scholars Ma Guohan 馬國翰 (1794-1857), Wang Mo 王謨 (c. 1731-1817) and Hong Yixuan 洪頤煊 (1765-1837). Ma's collection, included in his series Yuhan shanfang jiyi shu 玉函山房輯佚書, was based on Zhu Yizun's 朱彝尊 (1629-1709) book Jingyikao 經義考. There are two texts preserved in Japan that include two more fragments, namely the Bifulüe 秘府略, a text from the Tang period, and the Japanese text Shaku Benshō bunkyō hifu ron 釋遍照文鏡秘府論.
Among the texts discovered in the early Han-period tomb of Mawangdui 馬王堆 near Changsha 長沙, Huhan, in 1973 the Zhouyi 周易 was found. The text of the hexagrams Qin 欽 and Lin 林 in this version are very similar to surviving fragments of the Guicang, which might be a prove that the Guicang was at least already existing during the late Warring States period 戰國 (5th cent.-221 BCE), and not a forgery of Han-period persons.