The length of Spielberger's State Anxiety Inventory (SAI) makes its use difficult for critically ill populations. To shorten it, exploratory factor analysis was used with data from 530 acute myocardial infarction patients. The resulting 6-item shortened version demonstrated good internal reliability consistency with a Cronbach's alpha of .84. Construct validity of the shortened version was supported by strong correlations with the original SAI and the Brief Symptom Inventory (BSI), another measure of state anxiety, and through support of known hypotheses. This study offered a short version of the SAI with excellent psychometric properties that can be used in critically ill patients.