In recent years, NHS Ambulance Trusts throughout the UK have consistently failed to achieve their response time targets for both actual and potential life-threatening calls. To avoid a media and public outcry, the NHS response has been to change the basic parameters upon which the response time targets are calculated. An action learning study, which considered patient experience from initial response to outcome, concluded that the ambulance service must move away from the nearest crew response model to one which provides a defined multi-organisational service to specific categories of need. A key issue with the learning sets, which were made up of front-line crews, was understanding the Trust-wide picture and where acceptable new procedures could provide economic benefits to the Trust, benefits to the patients and help achieve the response time targets. A simulation model driven by parameters agreed by the action learning sets provided proof that new procedures would generate the required benefits. The learning sets also identified that the public should adjust their expectations to understanding that an immediate front-line ambulance response would only be despatched in life-threatening cases, but there would be alternative slower responses for all other cases.