Housing for migrant workers has become a major social challenge to the large-scale urbanization in China. Many studies have demonstrated housing disadvantages of temporary migrants in thatmigrants have limited access to both market-based housing due to their economic conditions and government-
sponsored affordable housing due to their official residency (hukou). In particular，much literature has analyzed institutional and individual factors of migrant housing conditions. This paper，however，investigates the extent to which migrant workers are able to move from employer-provided dorms or housing provided by social ties to market-based housing while working and living in cities，and institutional and individual factors contributing to such changes. Statistical analysis using data derived from a twelve-city migrant survey shows that，whereas most migrant workers live in dorms from employers and housing from social ties upon moving to cities，many of them gradually move to rent or even purchasing housing from formal and informal housing markets. The odds of such change are significantly influenced by factors such as years of working in the city，types of migration，willingness to permanently settle in the city，migrant social capital，education and types of Danwei .  

Drawing upon data from a twelve-city migrant survey conducted in 2009， this paper takes the advantage of structural equation modeling to examine the determinants of settlement intention of migrant workers， and to explore the mechanism of how these determinants indirectly affect settlement intentions via
city attachment． Model results show that the sense of city attachment plays an important role in determining rural migrantsintention to make permanent settlement in cities． Migrants of more years of education，being single， having skilled job， and with spouse or boy or girl friend in the same city are more willing to settle in cities． Having access to formal housing， frequently interacting with local people， and mastering local language not only positively affect the intention to settle in cities， but also indirectly and positively affect settlement intention via city attachment． The direct negative effect of size of social network on settlement intention is greater than the indirect positive effect via city attachment． Experience of discrimination and gender role also indirectly affect settlement intention． Model results also demonstrate the significance of further distinguishing the direct and indirect effects．