The research builds on the knowledge that under high-rocketing housing prices in many attractive cities with favorable em-ployment prospects, some talented workers are forced to leave due to poor housing affordability. To retain them, some municipalities adopt housing subsidies for targeted talents as a policy instrument. In China, such policies are named as Talent Housing (TH) policies. However, due to complex effects (i.e., positive, negative, composite) of housing prices on talent flow, along with impacts of cities’ socio-economic conditions on talent flow, policymakers are uncertain about whether TH policies will have expected outcomes. Therefore, the research aims to explore the impact of TH policies on talent agglomeration, along with its relevant determinants in terms of cities’ economic conditions (e.g., work opportunities and housing affordability). Generalized Method of Moment (GMM) and Finite Distributed Lag Model (FDLM) were applied to analyze the causality between talented human capital and TH subsidies of 70 Chinese cities. The findings show that TH subsidies are conducive for talent agglomeration, in particular to home-buying subsidies which have a 1-year lag positive effect. Nevertheless, TH subsidies might be less effective in cities with poor work opportunities and high housing affordability. The research provides insights into the economic impacts of TH policies on talent flow and helps policymakers from different cities design more suitable TH policies for sustainable development.