In the last decades, many parts of the world faced an increase in the number of extreme weather events and worsening climatic conditions with negative impacts for local populations and their livelihoods. While various empirical studies have identified key factors of disaster preparedness and vulnerability, we still lack a conceptual understanding of how these forces interact and how they impact household decision making. This study develops a dynamic household model, in which households face stochastic environmental hazards, which can lead to a loss of their wealth. To respond to the risk, households can either relocate to a safer area or undertake preventive measures to protect their physical assets. Both actions require material and immaterial resources, which constrain the household’s decision. Households are assumed to be heterogeneous with respect to key empirically identified factors for individual disaster risk: education, income, risk awareness, time preference and their access to preventive measures. Theoretical insights on the optimal household strategies are derived from the FOCs and the HJB equations. Furthermore a numerical solution for the optimal policy functions is presented calibrated to data from Thailand and Vietnam. Using Monte-Carlo-Simulations the corresponding stationary distributions are derived and compared to their empirical counterparts and the impacts of different household characteristics with a special focus on education are systematically assessed.