Abstract: A fish fossil collected in 1997 from Huachi area, Gansu province is described in this paper. Preliminary study has suggested that the fossil could be identified as the coelacanthid, which previously was only reported from a few localities in China. Description The specimen is very incomplete, the head, the dorsal part of the body and the middle lobe of the caudal fin were worn away by drilling. However, the other parts are better preserved, therefore, many features of the specimen could stll1 be observed. The scales are small, very thin and likely rhombic though their outlines are difficult to distinguish. The surface of the scales is ornamented with striking strike running mainly rostrocaudaly and a few striae are converged anteriorly. In the anterior division of the vertebral column, the neural arches are poorly preserved, but int he caudal region they are well preserved and characterized by a broad basal part and a long spine connected distally with the proximal end of radial of the caudal fin. The haemal arches are well developed in the caudal regions, they are higher fork-shape in the proximal part and produced into a long spine in the distal part which is connected distally with the proximal end of the radial of the caudal fin. The paired, anal and caudal fins are not preserved perfectly, but some features of them are visible.The pectoral is smaller than the pelvic, situated slightly higher and composed of about 17 slender rays. The pelvic is very far forwards and lies slightly behind the pectoral. Therefore, the pelvic is obviously thoracic and made up of about13 rays that are smooth and distally jointed. The anal fin is small, consisting of about 15 rays. The caudal fin is more completely preserved except the middle lobe and major features call be easily distinguished. The fin is fairly long mad distinctly diphycerca1.The dorsal and ventral lobes are not abruptly truncated instead, they decrease posteriorly and gradually in height. The backbone runs straight to the tip of the tail, and the fin is arranged symmetrically above and below it. Comparison The features described above clearly suggest that the specimen from Huachi area. Gansu province could be assigned to the order Coelacanthiformes. Berg(1955)subdivided this order into three suborders, namely Diplocercioidei, Coelacanthoidei and Laugioidei. Since the pelvic fin of Huachi coelacanth is characterized by the thoracic position, it should he referred to the suborder Laugioidei. According to Berg (1955), the suborder Langioidei consists only one family Laugiidae comprising only the type genus Laugia, However, Romer (1966)and Forey (1998) have different opinion on the classification. They assigned respectively Holophagus and Coccoderma to the family Laugiidae. Nevertheless, the coelacanth from Huachi differs apparently from Hotophagus in that the pelvic fin is situated very far forwards and the scales ornamented with striking striae running mainly rostrocaudally. The materials described under the generic name of Coccoderrma are mainly somee bone plates of the head and can not be compared to the present specimen. Our specimen is similar somewhat to Laugia in such features as the caudal and scales, as well as the thoracic potation of the pelvic fin, but it is different from Laugia because the pelvic rays are very slender. Due to its incomplete preservation, we could not make a comprehensive comparison between our specimen and Laugia. Therefore, we are not certain whether the specimen described here is really belonged to the genus Laugia or represents a new form of the family Laugiidae. Here in this paper, we tentatively refer our specimen to an undetermined species and genus of Laugiidae. Discussion During the evolutionary history of the coelacanths, the early members of the group were generally fresh water types. However, they have been found in a number of marine deposits in the Mesozoic. We believe that the coelacanth-bearing layer in Huachi area, Gansu province is probably related to the marine environment. Horizon The Upper Triassic Yanchang Group