The rapid socio-economic development in China has given rise to the problem of housing obsolescence. The reduction of housing obsolescence and the increase in housing service life are important from the perspective of sustainable development. In extant literature, flexibility has been introduced as an important method to reduce housing obsolescence. However, no empirical evidence exists to illustrate the effects of flexibility on reducing housing obsolescence. Therefore, by correlating knowledge on flexibility and housing obsolescence, this study attempts to explore the effects of flexibility on reducing urban housing obsolescence with special focus on Chinese flexible housing practices during the mid-1980s to the early 1990s. The research question which this study attempts to address is whether housing obsolescence can be obviously reduced by flexibility in Chinese flexible housing practices during the mid-1980s to the early 1990s. First, the theoretical framework of this research explains Chinese flexible approaches within the context of Chinese housing development since 1949. Second, it explores the mechanism of obsolescence in flexible housing practices during the mid-1980s to the early 1990s. The internal and external threat of housing obsolescence is discussed. A cross-national comparison is also made to understand the context of these practices. Third, theoretical framework suggests two scenarios and three theoretical propositions, and it specifies three key factors for analyzing the effects of flexibility on the reduction of housing obsolescence: changeability, housing obsolescence, and changing behavior by dwellers. Finally, it establishes a methodology for empirical investigation. Six Chinese neighborhoods built during the mid-1980s to the early 1990s which contain both flexible houses and similar ordinary houses are selected as cases according to screening criteria. Data are obtained by survey among households, interviews, analysis of archival records, and on-site observation. The main unit of analysis is a type of flat. The data analysis revolves around the three propositions put forward. The results of the empirical study indicate that the six cases do not agree with the three propositions. First, only a small number of households use the pre-set changeability in flexible flats. Second, flexible flats do not have a much lower score in housing obsolescence than similar ordinary flats. Third, the score on housing obsolescence for households who use changeability is not obviously lower than that for households who do not. These findings are discussed within the theoretical framework. In general, this research does not support the notion that flexibility in the flexible housing can reduce housing obsolescence obviously in the given context. ; published_or_final_version ; Architecture ; Doctoral ; Doctor of Philosophy