Patients with Alzheimer's disease participated in a trial of two sessions in which they received physostigmine and neostigmine in a double-blind crossover design. Most of these patients subsequently participated in a scopolamine vs saline double-blind crossover trial using a similar design. Physostigmine increased plasma cortisol relative to neostigmine, with the greatest difference at time points greater than 90 min post drug oral administration. Physostigmine also significantly decreased plasma cholinesterase (ChE). There was a significant positive correlation between the effects of physostigmine on increasing cortisol and decreasing ChE; there was no correlation between the increase in cortisol of cholinesterase inhibitor following neostigmine administration, but neither of these chemical parameters is related to the drug's effects on cognitive functioning.