OBJECTIVE: To determine the relationship of electrocardiographic rhythm during cardiac arrest with survival outcomes. DESIGN: Prospective, observational study. SETTING: Total of 411 hospitals in the National Registry of Cardiopulmonary Resuscitation. PATIENTS: Total of 51,919 adult patients with pulseless cardiac arrests from April 1999 to July 2005. MEASUREMENTS AND MAIN RESULTS: Registry data collected included first documented rhythm, patient demographics, pre-event data, event data, and survival and neurologic outcome data. Of 51,919 indexed cardiac arrests, first documented pulseless rhythm was ventricular tachycardia (VT) in 3810 (7%), ventricular fibrillation (VF) in 8718 (17%), pulseless electrical activity (PEA) in 19,262 (37%) and asystole 20,129 (39%). Subsequent VT/VF (that is, VT or VF occurring during resuscitation for PEA or asystole) occurred in 5154 (27%), with first documented rhythm of PEA and 4988 (25%) with asystole. Survival to hospital discharge rate was not different between those with first documented VF and VT (37% each, adjusted odds ratio [OR]) 1.08; 95% confidence interval [CI] 0.95-1.23). Survival to hospital discharge was slightly more likely after PEA than asystole (12% vs. 11%, adjusted OR 1.1; 95% CI 1.00-1.18), Survival to discharge was substantially more likely after first documented VT/VF than PEA/asystole (adjusted OR 1.68; 95% CI 1.55-1.82). Survival to discharge was also more likely after PEA/asystole without subsequent VT/VF compared with PEA/asystole with subsequent VT/VF (14% vs. 7% for PEA without vs. with subsequent VT/VF; 12% vs. 8% for asystole without vs. with subsequent VT/VF; adjusted OR 1.60; 95% CI, 1.44-1.80). CONCLUSIONS: Survival to hospital discharge was substantially more likely when the first documented rhythm was shockable rather than nonshockable, and slightly more likely after PEA than asystole. Survival to hospital discharge was less likely following PEA/asystole with subsequent VT/VF compared to PEA/asystole without subsequent VT/VF.