This article presents an evaluation of a rapid-response crisis-intervention service, hospice at home (H@H), for patients with advanced cancer. The project took place in Glasgow, UK, between July 1999 and May 2001. An integral part of the (H@H) project was the concurrent evaluation, which attempted to explore a range of service and user outcomes. The service was able to prevent admission to, or facilitate discharge from, institutional care on 62 occasions. The evaluation found significant improvements in some areas of pain and symptom management. High levels of satisfaction were recorded by all service users. A partial cost analysis revealed that the medical and nursing support costs for the (H@H) would have been substantially reduced if throughput had been higher. The (H@H) project team agreed that a valuable lesson learned from the project was the importance of involving all key players from the outset when determining the requirements of a new service initiative.