The current increase in the cost of health care must be considered as a severe threat to the prehospital emergency services system. Two examples have been selected--the patient with polytrauma and the patient in cardiac arrest--to demonstrate the dilemma between a need for objective data and the requirements of emergency patients. Study results obtained in trauma patients indicating that total prehospital time, including scene time, is correlated to patient outcome have led to the conclusion that at the scene treatment by emergency physicians may be dispensable. It has, however also been demonstrated that the time required for medical treatment at the scene is equivalent to 20% of the total scene time, thus representing only a fraction of the total prehospital time. Correlating the total prehospital time or scene time to outcome therefore appears absurd. The treatment principle of aggressive shock treatment in polytrauma needs critical reevaluation on the basis of results obtained by recent preclinical studies in patients with penetrating torso injuries. Small volume resuscitation could not be demonstrated to improve outcome in polytrauma patients, although a slight improvement in patients with brain injury may be assumed. Endotracheal intubation and early artificial ventilation are proven therapeutic principles in polytraumatized patients. Unfortunately, for ethical reasons randomised carefully controlled comparative studies can not be performed in polytrauma patients unless the patient is fully conscious. The importance of endotracheal intubation and artificial ventilation in unconscious trauma patients becomes apparent under conditions of anaesthesia where the application of the endotracheal tube averts regurgitation, aspiration and concomitant morbidity and mortality. The common causes of cardiovascular collapse and their pathomechanisms, as well as the mechanisms of cardiopulmonary resuscitation, have been widely investigated. Nevertheless, various aspects of their application are still controversial. The most recent study results have recommended initial ventilation prior to thoracic compression. New methods of assisting mechanical cardiopulmonary resuscitation, such as ACD CPR or vest CPR, have shown promising results in animal experiments. However, the importance of results obtained by preclinical randomised controlled investigations in humans need to be confirmed by further studies as to outcome. The efficacy of defibrillation in cases of ventricular fibrillation has been clearly demonstrated, particularly with a view to the interval between ventricular fibrillation and defibrillation. It has further been demonstrated that basic cardiopulmonary resuscitation preserves ventricular fibrillation and thus improves the chance of survival. The present generation of defibrillators has been further improved, particularly by the introduction of biphasic defibrillator wave forms, which may reduce the required energy, as well as possible complications, while offering an increase in the efficacy of defibrillation and a reduction in defibrillator size. Scientific emergency medicine is responsible not only for the development and validation of new methods and concepts, but in particular for their application under quality control conditions. Politicians require an improvement in the quality of the validation of emergency measures, although the instruments available for the investigation of these measures are known to be obsolete (experimental models, experimental design). Additionally, the financial support of research in emergency medicine suffers from being accourded low priority by public research funds such as the German Research Fund. However, in view of the rapid application of experimental results to daily practice it should be emphasized that patients also support research in emergency medicine via their direct financial contributions to the health insurance companies.