To reduce the operational costs and mitigate the risks of shortage, this paper develops a decision tool to determine the optimal facility location, quantities and inventory of multiple relief supplies in disaster operation management. Considering the uncertainty of demand for relief supplies, a two-stage stochastic programming model is proposed with the objective of minimizing fixed costs, procurement costs, transportation costs, holding costs and penalty costs. The Conditional Value-at-Risk (CVaR) is employed to measure penalty costs induced by shortage of supply. A Sample Average Approximation (SAA) approach is adopted to solve reasonably large-scale problems. The testing data setting is generated for numerical experiments. The applicability of the proposed model is studied, whilst the effectiveness of the proposed SAA is illustrated.