Evaluation of environments in hospital settings with regard to their therapeutic qualities has been a problem for clinicians and evaluation specialists for sometime. The present project examined the Ward Atmosphere Scale (WAS) factor-analytically and derived second-order factors inconsistent with those proposed by the authors of the scale. Items of the WAS were arranged psychometrically into three scales as implied by the factor analysis. Results suggest that interpretation of the WAS can be simplified considerably if these three super-ordinate dimensions are used to organize item-level data.