Cognitive ratings that use bipolar constructs based upon similarity and contrast have been shown to be biased towards the similarity pole in approximately a 62/38 ratio. This bias has also been known to shift in the contrastive direction for individuals who have psychiatric problems. This quantitative measure of cognitive change has a potential for characterizing cognitive changes that occur during the disease process, including recovery from disease. The present study investigated changes in self-aspect ratings and bipolar construct use in adult male veterans who had undergone coronary artery bypass graft surgery. Results indicated that treatment subjects' self-aspect and construct ratings were more negative than controls'. Results also indicated that all subjects rated core interpersonal self-aspects closest to the expected bias, while self-aspects related to cardiac recovery problems were rated in the most contrastive direction. The results finally suggested that the greatest degree of change for the treatment subjects were in emotionally generated constructs. The results suggested a preliminary validation for characterizing cognitive changes in the disease process by measuring shifts in bipolar construct ratings.