Quantified sensitivities of heating and cooling loads to different variables that influence heat gain and loss in a building provides a valuable basis for energy efficient design, especially in temperate climate zones where particular parameter settings could be beneficial in one season while reducing performance or neutral in the other. In doing so it is important in this multi-parameter design space to consider impact of changes in each parameter when other variables also change. Such 2-variable up to n-variable correlation is called factorial analysis. The methodology is introduced using three variables (roof solar absorptance, air exchange rates, and sub-roof R-value) in a simple structure with all other parameters fixed. Sensitivity is via impact of changes on each of heating load, cooling load and annual total. Knowledge of factorial effects is shown to be important and lead to simple strategies that provide large benefits in both seasons. They also show that some standard approaches to saving energy (e.g. raising R significantly), while useful are often unnecessary, unless poor settings are made in other parameters. (C) 2012 Elsevier B.V. All rights reserved.