Shang is the earliest state to have left written records in Chinese history. Since the 1930s, historians and archaeologists have devoted great efforts to the study of its social organization. In retrospect, however, we saw that in previous studies energy was mainly invested in labeling society with theoretical jargon. This meant that the detailed social structure of the state remained poorly understood. There has been a rapid increase in the quantity of Shang mortuary data from excavations over the past few decades. By 2003 more than 10,000 Shang tombs had been excavated at Yinxu, the last capital city of Shang. These data contain a considerable amount of information on Shang society. Using this wealth of mortuary data from Yinxu, this thesis research is a study of Shang social organization from a mortuary perspective. Following theoretical and methodological developments in mortuary archaeology, I have analyzed more than 2000 burials at Yinxu. This study reveals that these burials are distributed hierarchically in clusters at three different levels that are labeled A-, B- and C-level clusters from lower to higher levels. Most A-level clusters are found to contain one set of paired-tombs. By analyzing the patterns of paired burials and sex-relations of the dead, it may be inferred that the dead in the paired burials within each A- level cluster were the members of the same family, possibly married couples. On the basis of the discussion of the paired-tombs, I argue that each A-level cluster might represent a nuclear family and each B-level cluster is composed of one or more A-level clusters; and it, in many cases, might represent an extended family. Each C-level cluster consists of a number of B-level clusters; and it may be interpreted as representing a single lineage. Within a C-level cluster is usually a group of large tombs that might belong to lineage chiefs. It is, therefore, suggested that the Shang society might be organized horizontally in lineages. Available mortuary data in this study also provide significant amount of information on vertical social organization. The Shang was a highly stratified society; and it was structured with different social classes from highest to lowest in terms of their social status. Based on the burial database used in this study, it is estimated that the social group including lineage chiefs and those of higher classes counted less than 1% of the total population, approximately the class of lineage elites 7-10 %, commoners 82- 87%, and lowest class 3-7%. The high proportion of commoners in the total population gives no support to the long-held proposition that Shang was a slave-based society. This study shows considerable variability among different lineages in terms of their social status as reflected in mortuary practices. It also suggests that males did have a higher social status than females.