Older people with dementia living in residential care have complex unmet needs and decreased quality of life. To reduce unmet needs in older people with dementia in residential care compared to a 'care as usual' control group. A single blind, multicentre, cluster randomised controlled trial recruited 238 people aged 60+ with dementia living in 24 residential homes from three areas. Unmet needs were measured using the Camberwell Assessment of Needs for the Elderly (CANE) and quality of life using the Quality of Life in Alzheimer's Disease (QoL-AD). Homes were randomised to the control (care as usual) or the intervention group 1 hour per week liaison input per home to deliver a personalised intervention package over a 20 weeks to meet the unmet needs. A single blind follow-up included 192 (81%) available participants. At follow-up the total number of unmet needs was reduced in both the intervention and control groups, but analysing the groups by clusters there were no significant differences in either unmet needs or quality of life. The unmet needs of people with dementia can be identified using the CANE. The CANE assessment may have led to unmet needs being reduced at follow up, but the liaison intervention did not significantly reduce total unmet needs relative to the control group. Unmet needs such as sensory problems, mobility, drugs, and psychological distress were especially reduced in the intervention group at follow up. Copyright 2007 John Wiley & Sons, Ltd.