Modern product design cycles are requiring faster turnaround of design changes and the accompanying noise, vibration and harshness analyses. Design sensitivity analysis (DSA) guides the designer in making changes by indicating the model variables that cause the greatest benefit in noise or weight reduction. The use of analytical models based on statistical energy analysis (SEA) is attractive for predicting noise and vibration environments because of their simplicity and solution speed compared with deterministic models. This paper describes the implementation of DSA in an SEA computer code. Examples of SEA/DSA for controlling airborne and structureborne noise problems are described.