Abstract: Whether panda belongs to Ursidae,Procyonidae or an independent family had been debated until the 1980s,when O Brien and others posed a series of researches on the protein and DNA of the giant panda.Until the 1990s,most of the western scholars agreed that panda belonged to Ursidae while most of the Chinese scholars insisted that it should be an independent family,considering every factor of biology and researches on the molecular level as well as on palaeontology.QIU Zhan-xiang et al.studied the fossils of pandas ancestors found in Yunnan Province in 1989.HUANG Wan-bo studied pandas skulls,mandibles and teeth in 1993.All these studies have further proved that it was in the later Miocene period that the Ailurarctos and Ursavus began to develop parallelly:the former had developed to panda while the latter had developed to bear (all the species in Ursidae). The studies on population ecology of panda during the latest years are summarized overall in five aspects.The first is population.With different methods,we believe that there is 200 or so in Qingling mountains,300 or so in Mingshan and Qionglai mountains,less than 40 in Daxiangling and Xiaoxiangling mountains,more than 100 in Liangshan mountains,the total is 1000 or so.The second is population characteristic.The third is population growth and fluctuation.The fourth is genetic diversity.The genetic diversity from high to low in different mountains is:Qionglaishan→Mingshan→Qingling→Liangshan→Xiaoxiangling→Daxiangling.There is distinct genetic differentiation in different mountains,especially in polymorphic aspect of genetic group.There are three different genetic groups:the populations in Qingling mountains,Mingshan-Qionglai mountains and Liangshan mountains are in kinship.There are close kinship between the populations in Mingshan-Qionglai mountains and Liangshan mountains,but their kinship with Qingling mountains is quite far.The fifth is viability analysis.At the same time,some problems existed among them are brought up and discussed.