A palliative care team in north Scotland identified serious drug errors occurring in the local health-care trust. These were connected with the use of variable rate syringe drivers (IVAC P100) to deliver pain and symptom management rather than the more suitable Graseby MS26. An action research approach was taken to effecting change. An educational workshop was set up which 23 nurses attended. Clinical support was provided and 3 months later an evaluation was carried out. Of the 13 nurses who had used an MS26, most were able to correctly follow the process of setting up the driver and had made appropriate observations, but drug calculations were a problem. Change was stated to have taken place in seven clinical areas but the degree of change was variable. Facilitators and obstacles to the change process are identified and recommendations made for the next round in the action research cycle.