This study evaluated the efficacy and safety of donepezil in patients with mild to moderately severe Alzheimer's disease, and examined the relationships between plasma donepezil concentration, red blood cell acetylcholinesterase (AChE) activity and clinical response. The trial was of a multicenter, double-blind, parallel-group design and patients were randomised to once-daily treatment with either donepezil (1, 3 or 5 mg) or placebo. The 12-week double-blind phase was followed by a 2-week single-blind placebo washout. 161 patients (55-85 years of age) entered the study and 141 completed treatment. Patients treated with donepezil showed dose-related improvements in the Alzheimer's Disease Assessment Scale-cognitive subscale score (ADAS-cog) and in MMSF scores. The improvements in ADAS-cog were statistically significantly greater with donepezil 5 mg/day than with placebo. There was a 50% reduction in the percentage of patients showing clinical decline with donepezil at 5 mg/day (11%) relative to placebo (20%). In addition, a statistically significant correlation between plasma concentrations of donepezil and AChE inhibition was demonstrated. A plateau of inhibition (76-84%) was reached at plasma donepezil concentrations > 50 ng/ml. The correlation between plasma drug concentrations and ADAS-cog (p = 0.014), MMSE (p = 0.023) and patient quality of life scores, assessed by the patient (p = 0.037) were also statistically significant, as was the correlation between AChE inhibition and change in ADAS-cog (p = 0.008). The incidence of treatment-emergent adverse events with all three dosages of donepezil (64-68%) was comparable to that observed with placebo (65%). Donepezil had no clinically significant effect on vital signs, haematology or clinical biochemistry tests. Importantly, donepezil was not associated with any hepatotoxicity, as observed with acridine-based cholinesterase inhibitors.