BackgroundA recent Cochrane review of reality orientation therapy identified the need for large, well-designed, multi-centre trials.AimsTo test the hypothesis that cognitive stimulation therapy (CST) for older people with dementia would benefit cognition and quality of life.MethodA single-blind, multi-centre, randomised controlled trial recruited 201 older people with dementia. The main outcome measures were change in cognitive function and quality of life. An intention-to-treat analysis used analysis of covariance to control for potential variability in baseline measures.ResultsOne hundred and fifteen people were randomised within centres to the intervention group and 86 to the control group. At follow-up the intervention group had significantly improved relative to the control group on the Mini-Mental State Examination (P=0.044), the Alzheimer's Disease Assessment Scale – Cognition (ADAS–Cog) (P=0.014) and Quality of Life – Alzheimer's Disease scales (P=0.028). Using criteria of 4 points or more improvement on the ADAS–Cog the number needed to treat was 6 for the intervention group.ConclusionThe results compare favourably with trials of drugs for dementia. CST groups may have worthwhile benefits for many people with dementia.