This paper presents a technique used to examine the sensitivity of the output of a building energy model with respect to the variation of different design variables. The Monte Carlo-based sensitivity analysis was applied and a case-study house was used to demonstrate this technique. |