Abstract: The present invention is a method of diagnosing the presence of a persistent occlusion in a myocardial infarct patient undergoing thrombolytic therapy. The method comprises detecting a series of three variables from the patient and then generating the probability of the presence of a persistent occlusion from those variables. The first variable comprises a serum creatine kinase MB (CK-MB) level from a patient at the onset of thrombolytic therapy. The second variable comprises a second CK-MB level in the patient at a predetermined time after the onset of thrombolytic therapy. The third variable comprises the presence or absence of chest pain a predetermined time after the onset of thrombolytic therapy. In a preferred embodiment of the present invention, the second and third variables are detected within 30 minutes of each other and within about 1 to about 3 hours after the initial variable is detected.