Personal information security plays fundamental and critical role in promotion of smart cities. By taking personal information, vulnerability and threat as basic elements for risk assessment, this article proposes a Markov method-based personal information security risk assessment model in smart cities with the core of threats (Li Hetian, 2007). Based on threat probability, threat consequence attribute and attribute value acquired through the Markov method, threat analysis, the multi-attribute decisionmaking theory and the expert grading method, this article calculates the objective threat indexes, which is then utilized for risk ranking, so as to provide scientific basis for formulating targeted personal information security risk management and control strategies.