We aimed to describe the epidemiology of Rapid Response Team (RRT) activation in New Zealand public hospitals. We undertook a prospective multicentre observational study of RRT activations in 11 hospitals for consecutive 14-day periods during October–December 2014. A standardised case report form was used to collect data on patient demographics, RRT activation criteria and timing, vital signs on RRT arrival, team composition and intervention, treatment limitation and patient outcome at day 30. Three hundred and thirteen patients received 351 RRT calls during the study period. Patients were admitted under a medical specialty in 177 (56.5%) instances. Median duration from hospital admission to first RRT call was two days. Eighty-six percent of RRT calls were to inpatient wards. A total of 43.4% of RRT calls occurred between 0800 and 1700 hours (38% of the day) and 75.5% of RRT calls were activated by ward nurses. A median of three staff attended each call. Common triggers for RRT activation were increased Early Warning Score (56.2%) and staff concern (25.7%). During the RRT call, 2.8% of patients died; 19.8% died by day 30. New ‘Not For Resuscitation’ orders were written in 22.5% of RRT calls. By day 30, 56.2% of patients had been discharged home alive. In conclusion, RRTs in New Zealand are multidisciplinary, mostly nurse-activated and predominantly respond to deteriorating medical (rather than surgical) patients. Most patients remain on the ward. The RRT frequently implements treatment limitations. Given almost one in five patients die within 30 days, over half of whom die within 72 hours of RRT review, surviving the RRT call may provide false reassurance that the patient will subsequently do well.