The outcome of out-of-hospital cardiac arrest is very much determined by uncontrollable precardiopulmonary resuscitation (CPR) conditions. Two consecutively registered databases containing variables related to pre-arrest, arrest and CPR are similarly analysed to produce and validate a simple clinical algorithm for acute decision making during CPR. The outcome results in the two different time periods remained nearly unchanged. The simultaneous and persistent absence of ventricular fibrillation, gasping and light-reactive pupils after arrival of the second tier was strongly associated with a poor outcome. Unresponsiveness of these variables to a full and optimal trial of advanced life support can in itself be considered as an index for irreversible myocardial and neurological damage.