Objective: To evaluate the outcome and the factors concerned with of out-of-hospital cardiac arrest patients according to the location of the collapse. Methods: From May 1st, 1998 to April 30th, 2001, 15,211 consecutive out-of-hospital cardiac arrest cases considered for resuscitation were recorded. Of these cases 7540 arrests in subjects aged 18 years or older with cardiac aetiology were analyzed. The outcome and the related-factors, particularly incidence of ventricular fibrillation, were evaluated according to the location of the cardiac arrest. To analyze the factors that affect the incidence of ventricular fibrillation, a logistic regression model was used. Results: About three-quarters of out-of-hospital cardiac arrests occurred at private residences. The outcome and characteristics were significantly different according to the location of the arrest. Arrest patients in public or in the work place had a higher chance of being found in ventricular fibrillation and survival than those at a private residence. The multivariate adjusted odds ratios for ventricular fibrillation in a public or work place were significantly higher than that in private residences, after adjusting for covariates affecting initial rhythm, such as age, sex, witnessed status, bystander cardiopulmonary resuscitation, and response interval. Conclusion: Although the majority of out-of-hospital cardiac arrests occur at private residences, arrests in public or in the work place had a higher chance of being found in ventricular fibrillation and survival than those at private residences. In order to establish a system to improve the outcome of out-of-hospital cardiac arrest, a well-considered strategy considering the location of arrest is necessary. © 2005 Elsevier Ireland Ltd. All rights reserved.