The aim of the study was to develop a scoring system for outcome classification at the start of prehospital first tier resuscitation for patients with cardiac arrest from ventricular fibrillation (VF). We studied a consecutive sample of 100 out-of-hospital cardiac arrest patients, presenting with VF of presumed cardiac etiology on arrival of the first tier (in a two-tiered urban Emergency Medical Services system). The number of patients discharged was 29 ('survivors') and 71 died ('non-survivors'). The electrocardiography (ECG) tracings recorded during resuscitation using a semi-automatic defibrillator were retrospectively analysed. For each patient, VF amplitude in mV (VF_a) and the number of base-line crossings per second (VF_blc) were calculated. Fisher's linear discriminant analysis was applied to discriminate between survivors and non-survivors using the variables VF_a, VF_blc and age. Patients were classed as potential survivors or non-survivors using a survival index = 0.6*(VF_a) + 0.4*(VF_blc)-4.0. If for a given patient the survival index is < 0, he is classified in the non-survivor group, if the survival index is > 0, he is classified in the survivor group. Using this index 79% of the survivors and 70% of the non-survivors could be classified correctly. Adding age to the formula increased the correct classification of survivors to 86 and 73% for the non-survivors. The survival index provides a research tool for the discrimination between potential survivors and non-survivors, which opens the possibility for the development of alternative treatment protocols in cardiac arrest.