AIM: To describe the characteristics and outcome among patients suffering in-hospital cardiac arrest in relation to whether the arrest took place in a ward with monitoring facilities. METHODS: All patients who suffered an in-hospital cardiac arrest during a 4-year period in Sahlgrenska Hospital, Göteborg, Sweden and in whom resuscitative efforts were attempted, were prospectively recorded and described in terms of characteristics and outcome. RESULTS: Among 557 patients, 292 (53%) had a cardiac arrest in wards with monitoring facilities. Those in a monitored location more frequently had a confirmed or possible acute myocardial infarction (AMI) as judged to be the cause of arrest (P < 0.0001), and the arrest was witnessed more frequently (96 vs. 79%; P < 0.0001). Ventricular fibrillation/tachycardia was observed more often as initial arrhythmia in monitored wards (56 vs. 44%; P = 0.006). The median interval between collapse and first defibrillation was 1 min in monitored wards and 5 min in non-monitored wards (P < 0.0001). Among patients with arrest in monitored wards 43.2% were discharged alive compared with 31.1% of patients in non-monitored wards (P = 0.004). Cerebral performance category (CPC-score) at discharge was somewhat better among survivors in monitored wards. CONCLUSION: In a Swedish University Hospital 47% of in-hospital cardiac arrests in which resuscitation was attempted took place in wards without monitoring facilities. These patients differed markedly from those having arrest in wards with monitoring facilities in terms of characteristics, interval to defibrillation and outcome. A shortening of the interval between collapse and defibrillation in these patients might increase survival even further.