Background: Results of epidemiological studies, neuropathological observations, and in-vitro experiments all suggest that inflammatory mechanisms contribute to the destructive lesions in Alzheimer's disease. We aimed to establish the effect of the anti-inflammatory drug hydroxychloroquine on the progression of dementia. Methods: We did a double-blind, parallel-group, multicentre trial in which we randomly assigned 168 patients with early Alzheimer's disease to hydroxychloroquine (200 or 400 mg dependent on bodyweight), or placebo for 18 months. Outcome measures were related to activities of daily living, cognitive function, and behavioural abnormalities. Analysis was by intention to treat. Results: At 18 months, mean scores for the interview for deterioration in daily life in dementia in patients on hydroxychloroquine (22.6 [SD 11.4]) did not differ from those for patients on placebo (21.3 [10.5]). Also, mean scores on the cognitive subscale of the Alzheimer's disease assessment scale were closely similar in hydroxychloroquine (26.4 [14.9]) and placebo (25.7 [14.3]) treated patients, as were behavioural changes, measured by the revised memory and behavioural problems checklist (36.3 [12.0] and 34.2 [12.4], respectively). Explorative analyses did not suggest any specific subgroup that benefited from hydroxychloroquine. The frequency and nature of serious adverse events and side-effects were much the same in both groups. 155 (92%) patients completed all assessments over the entire study. Interpretation: Anti-inflammatory treatment with hydroxychloroquine for 18 months does not slow the rate of decline in minimal or mild Alzheimer's disease.