Seventeen normal volunteers received either 0.5 mg, 1.5 mg, or 2.5 mg physostigmine i.v. in a placebo-drug-placebo single-blind design. EEG was recorded simultaneously and analyzed by computerized spectral analysis. Eleven healthy elderly volunteers (mean age = 69.1 years) with mild memory impairment were treated with placebo, followed by oral choline chloride (either 8 g/day for 3 weeks, or 16 g/day for 1 week), and then, again, placebo. Recordings of spontaneous EEG and EEG event-related potentials (contingent negative variation) were obtained during both placebo and choline treatments. The larger doses of physostigmine produced an increase in low frequency activity and a slowing of the peak alpha frequency. Oral choline chloride had no effect on the EEG as measured by spectral analysis, but appears to have differential effects on contingent negative variation (CNV) amplitude and reaction time, depending upon the initial CNV amplitude.