Prevalence of bystander CPR and effect on outcome has been evaluated on 3053 out-of-hospital cardiac arrest (CA) events. Bystander CPR was performed in 33% of recorded cases (n = 998) by lay people in 406 cases (family members 178, other lay people 228) and by bystanding health care workers in 592 cases (nurses 86, doctors 506). Family members and lay people mainly applied CPR in younger CA victims at public places, roadside or at the working place. Sudden infant death syndrome (SIDS) and drowning are highly represented. Health care workers performed CPR mainly in older patients, at public places or at the roadside and especially in case of cardiac or respiratory origin. CA caused by trauma/exsanguination and intoxication/metabolic origin received less bystander CPR (23% resp. 22%). Cardiac arrests receiving bystander CPR are more frequently witnessed and have a shorter access time to the emergency medical service (EMS) system and shorter response time of basic life support (BLS). Advanced life support (ALS) response time is significantly longer. In witnessed arrests of cardiac origin receiving bystander CPR a significantly better late survival was observed. In non-witnessed arrests of cardiac origin early and late survival are significantly higher in patients receiving bystander CPR. In CA events where response time of ALS exceeds 8 min, the beneficial effect of bystander CPR is most significant. Furthermore no deleterious effect of bad technique or inefficient bystander CPR can be demonstrated.