The purpose was to explore the psychometrics and correlates of the Toronto Alexithymia Scale (TAS-20) after myocardial infarction (MI).The TAS-20 and other self-report measures were administered 3-6 months after discharge to 1443 patients.Good internal reliability was confirmed for the total TAS-20 and two subscales (F1 and F2). The F3 showed low internal consistency linked to negatively keyed items. The prevalence of alexithymia was 30.2% at the first interview. Alexithymics were older, less educated, more likely to have previous MIs and had higher scores on all measures of negative emotions. Six-month test-retest reliability was.47 (n=167). Residual change score analysis showed patients with more education and a first MI had greater decreases in alexithymia than expected.The TAS-20 has adequate internal consistency in post-MI patients, and its correlates are similar to other reports. Low temporal stability suggests that secondary alexithymia is important after MI.