Two competing hypotheses explaining gender bias in cardiac care were tested. The first posits that women's coronary heart disease (CHD) symptoms are simply misinterpreted or discounted. The second posits that women's CHD symptoms are misinterpreted when presented in the context of stress. In two studies, medical students and residents randomized to 2 (male vs. female) × 2 (stress vs. nostress) experiments read vignettes of patients with CHD symptoms and indicated their diagnosis, treatment, and symptom origin interpretation. Both studies disconfirmed the first hypothesis and strongly supported the second. Only when stress was added did women receive significantly lower CHD diagnoses and cardiologist referrals than men and did the origin interpretation of women's CHD symptoms (e.g., chest pain) shift from organic to psychogenic. Neither participants' gender nor their attitude toward women influenced assessments.