Abstract The results of over 130 chemical analyses of Chinese Greenware bodies are presented, obtained by atomic absorption spectrometry. A clear chemical difference is seen between the earlier Northern Zhejiang Yue wares and the later Longquan celadons. Interesting connections are observed between Northern Celadons (assumed to be from Yaozhou), Hangzhou Guan wares and early Jingdezhen Greenwares, which are tentatively explained by the flight of the Song court to Hangzhou in AD 1127, perhaps taking Imperial potters with them. The Longquan celadons contained a substantial number of kiln sherds, obtained by Nils Palmgren on his visit to the area in 1936. The availability of authentic kiln material is very unusual in Western studies of this kind, and particular attention is paid to these data. It was hoped to discriminate between the named kilns around Longquan (Lishui, Qikou, Yuankou, Longfeng, Jincun, Dayao, Anfu, Baoding and Wenzhou), and it proved possible to distinguish chemically the Baoding and Wenzhou products. These kilns are geographically slightly removed from Longquan itself, are probably earlier in date, and their composition appears to be intermediate between the Yue wares of Zhejiang province and the true Longquan celadons. The other kilns proved difficult to distinguish, and observed small differences are probably due to differences in date or fabric quality.